http://sumowiki.intec.ugent.be/api.php?action=feedcontributions&user=Icouckuy&feedformat=atomSUMOwiki - User contributions [en]2024-03-29T08:31:04ZUser contributionsMediaWiki 1.35.4http://sumowiki.intec.ugent.be/index.php?title=Useful_Links&diff=5835Useful Links2013-10-28T19:42:51Z<p>Icouckuy: /* Optimization */</p>
<hr />
<div>=== Related publications ===<br />
<br />
See the [[Related publications]] page.<br />
<br />
=== Related projects ===<br />
<br />
A list of projects with similar ideas and scope.<br />
* [http://www.cs.rtu.lv/jekabsons/regression.html] Matlab regression software<br />
* [http://sites.google.com/site/gptips4matlab GTIPS] Matlab symbolic regression software using genetic programming<br />
* [http://www.optiy.eu OptiY]<br />
* [http://www.nutechsolutions.com/prod_cv_analytics.asp ClearVu Analytics]<br />
* [https://sites.google.com/site/felipeacviana/ Surrogates Toolbox]: Another matlab surrogate modeling toolbox<br />
* [http://www.evolved-analytics.com Evolved-Analytics]: DataModeler package.<br />
* [http://www.muneda.com/ MunEDA]: MunEDA provides leading EDA software technology for analysis, modelling and optimization of yield and performance of analog, mixed-signal and digital designs.<br />
* [http://www.lsoptsupport.com/faqs/setting-parameters-for-metamodel-based-optimization-strategies LS-OPT: Functionality similar to the SUMO-Toolbox in LS-OPT]<br />
* [http://mucm.aston.ac.uk/MUCM/MUCMToolkit/index.php?page=MetaHomePage.html MUCH Toolkit]: A toolbox for sensitivity analysis using surrogate models<br />
* [http://home.mit.bme.hu/~kollar/topics/fdident.html FDIDENT: frequency domain identification toolbox]<br />
* [http://www.dmoz.org/Science/Math/Statistics/Software/Regression_and_Curve_Fitting/: Regression software on Open Directory]<br />
* [http://www.infiniscale.com/ Infiniscale]: TechModeler, TechAnalyzer: automatic model generation tools for EM applications<br />
* [http://www.vrand.com/visualDOC.html visualDOC]: VisualDOC, a design optimization tool<br />
* [http://www.salford-systems.com/mars.php MARS]: A spline based modeling tool<br />
* [http://www.phoenix-int.com/software/phx_modelcenter.php Modelcenter]: datamining and modeling tool<br />
* [http://www.simulia.com/products/isight2.html iSIGHT]: datamining and modeling tool<br />
* [http://www.tmpinc.com/datascape_overview.html Datascape]: datamining and modeling tool<br />
* [http://www.friendship-systems.com/ Friendship systems]: Tools for modeling ship hulls parametrically and performing the modeling calculations (Equilibrium tool)<br />
* [http://www.csse.monash.edu.au/%7Edavida/nimrod/nimrodg.htm Nimrod/G]: Execute parameter sweeps on the grid<br />
* [http://www-sop.inria.fr/oasis/ProActive/ ProActive]: Java grid library <br />
* [http://www.csse.monash.edu.au/%7Edavida/nimrod/nimrodo.htm Nimrod/O]: Grid-enabled optimization toolkit <br />
* [http://www.ece.northwestern.edu/OTC/ NEOS]: Distributed optimization toolkit <br />
* [http://software.sci.utah.edu/scirun.html SciRun]: Problem Solving Environment (PSE), for simulation, modeling, and visualization of scientific problems. <br />
* [http://www.esteco.it/ ModeFRONTIER]: environment dedicated to the set up of design assessment chains and efficient investigation of the design space. <br />
* [http://www.gridbus.org/ Gridbus]: metascheduler for the grid <br />
* [http://icl.cs.utk.edu/netsolve/overview/ Gridsolve/Netsolve]: grid enabled scientific computation toolbox <br />
* [http://www.dtreg.com/ DTREG]: a powerful statistical analysis program that generates classification and regression trees and Support Vector Machine models that can be used to predict parameter values. <br />
* [http://www.soton.ac.uk/%7Epbn/MDO/ MDO]: A collection of MDO links <br />
* [http://www.research.att.com/~njas/gosset/index.html GOSSET]: A general purpose program for designing experiments <br />
* [http://www.cs.sandia.gov/DAKOTA/ DAKOTA]: Design Analysis Kit for Optimization and Terascale Applications<br />
* [http://www.cs.cf.ac.uk/gridprojects/dipso/ DIPSO]: Wide-Area Distributed Problem Solving (DIPSO) <br />
* [http://www.geodise.org GEODISE]: Grid Enabled Optimisation and Design Search for Engineering<br />
* [http://www.fast.u-psud.fr/ezyfit/ EZfit] : Free curve fitting toolbox for matlab<br />
* [http://www.ians.uni-stuttgart.de/spinterp/about.html SGIT] : A Sparse grid interpolation toolbox<br />
* [http://www.csie.ntu.edu.tw/~yien/quickrbf/quickstart.php QuickRBF] : an RBF fitting library (native)<br />
* [http://www.farfieldtechnology.com/products/toolbox/ FastRBF] : another RBF fitting library (matlab)<br />
* [http://promethee.irsn.org PROMETHEE project: easy parametric modeling]<br />
* [http://simlab.jrc.ec.europa.eu/ Sensitivity Analysis library]<br />
* Gaussian Process Matlab code<br />
** [http://www.gaussianprocess.org/gpml/code/matlab/doc/ Code] based on Rasmussen's book<br />
** [http://www.cs.man.ac.uk/~neill/gp/ Gaussian Process Software]<br />
** [http://www.ios.htwg-konstanz.de/joomla_mof/index.php?option=com_content&view=article&id=48:polyreg-polynomial-gaussian-process-regression&catid=36:code&Itemid=81 GP] using polynomial covariance functions<br />
* [http://shark-project.sourceforge.net/ Shark Machine Learning Library - An open source machine learning bundle]<br />
<br />
=== Related labs ===<br />
<br />
A list of some of the labs/researchers with similar ideas and scope.<br />
<br />
* [http://aerospace.engin.umich.edu/index.html Department of Aerospace Engineering at the University of Michigan]<br />
* [http://www.mae.ufl.edu/~mdo/ The Structural and Multidisciplinary Optimization Group at the University of Florida]<br />
* [http://web.engr.oregonstate.edu/~tgd/ School of Electrical Engineering and Computer Science, Oregon State U]<br />
* [http://www.soton.ac.uk/engineering/research/groups/ced.page Southampton University Computational Engineering and Design Group]<br />
* [http://edog.mne.psu.edu/ Engineering Design & Optimization Group (Penn state)]<br />
* [http://nd.edu/~ame/ Aerospace and Mechanical Engineering] [http://www.gano.name/shawn/ Homepage]<br />
* [http://shyylab.engin.umich.edu/research/design-optimization Computational Thermo-Fluids Group]<br />
* [http://www.ensc.sfu.ca/~gwa5/index.htm Product Design and Optimization Laboratory (PDOL)]<br />
* [http://www.cerfacs.fr/4-25708-Home.php Computational Fluid Dynamics (CFD) group at CERFACS]<br />
* [http://webuser.uni-weimar.de/~roos1/ Dirk Roos]<br />
* [http://www.cerfacs.fr/~duchaine/HTML/research.htm Florent Duchaine]<br />
* http://www.nlr.nl/ National aerospace lab]<br />
<br />
=== Data sets - Simulation code ===<br />
<br />
A list of publicly available datasets and simulation codes, useful for testing.<br />
<br />
* [http://www.mat.univie.ac.at/~neum/stat.html Statistics links] : A nice collection of data fitting and analysis codes<br />
* [http://en.wikipedia.org/wiki/Classic_data_sets Wikipedia Classic Datasets]<br />
* [http://www.uni-koeln.de/themen/Statistik/data/rousseeuw/ The ROUSSEEUW datasets]<br />
* [ftp://ftp.sas.com/pub/neural/dojo/dojo.html Donoho-Johnstone Benchmarks] <br />
* [http://people.scs.fsu.edu/~burkardt/datasets/datasets.html datasets]<br />
* [http://www-syscom.univ-mlv.fr/~vignat/Signal/Space/index.html Spice like simulator for matlab]<br />
* [http://www.itl.nist.gov/div898/strd/general/dataarchive.html Nist dataset archive]<br />
* [http://homes.esat.kuleuven.be/~smc/daisy/daisydata.html Daisy datasets]<br />
* [http://lib.stat.cmu.edu/datasets/ Statlib dataset archive]<br />
* [http://www.iau.dtu.dk/nnbook/systems.html Datasets from the book "Neural networks for the modeling and control of dynamic systems.]<br />
* [http://www.gpc.de/e_poses.html A simulation environment for production and transport, logistics and automation systems]<br />
* [http://shumway.physics.asu.edu/codes.html Nanostructure Simulation and Modeling Programs]<br />
* [http://pedsim.silmaril.org/ A Modular, Distributed Pedestrian Crowd Simulation System]<br />
* [http://g95.sourceforge.net/g95_status.html Fortran simulation codes]<br />
* [http://www.genie.ac.uk/ Grid ENabled Integrated Earth system model]<br />
* [http://gcmd.nasa.gov/KeywordSearch/Home.do?Portal=GCMD&MetadataType=0 Nasa datasets]<br />
* [http://funapp.cs.bilkent.edu.tr/DataSets/ Function approximation repository]<br />
* [http://www.idsia.ch/~andrea/Andrea_Rizzoli_Home_Page/Sim_Tools.html A Collection of Modeling and Simulation Resources on the Internet]<br />
* [http://www.grc.nasa.gov/WWW/K-12/freesoftware_page.htm Free simulation software from Nasa]<br />
* [http://opensees.berkeley.edu/index.php Earthquake simulation]<br />
* [http://www.pdl.cmu.edu/DiskSim/ Harddisk simulator]<br />
* [http://spib.rice.edu/spib/mtn_top.html Mountain top radar data]<br />
* [http://www.statsci.org/datasets.html Statsci dataset repository]<br />
* [http://statwww.epfl.ch/davison/BMA/Data4BMA/ dataset repository]<br />
* [http://astrostatistics.psu.edu/datasets/ dataset repository]<br />
* [http://www.maths.uq.edu.au/CEToolBox/ problems from the Cross Entropy toolbox]<br />
* [http://www.cs.waikato.ac.nz/~ml/weka/index_datasets.html Weka datasets]<br />
* [http://www.ailab.si/orange/datasets.asp?Inst=on&Atts=on&Class=on&Values=on&Description=on&sort=Data+Set dataset repository]<br />
* [http://www.itee.uq.edu.au/%7Emarcusg/msg.html Max Set of Gaussians Landscape Generator]<br />
* [http://www.mathworks.fr/matlabcentral/fileexchange/loadAuthor.do?objectId=364966&objectType=author Collection of useful Matlab scripts by Thomas Abrahamsson ]<br />
* [http://www.mathworks.com/matlabcentral/fileexchange/loadFile.do?objectId=10731&objectType=FILE# A Matlab toolbox for Linear Structural Dynamics Analysis]<br />
* [http://webscripts.softpedia.com/script/Scientific-Engineering-Ruby/Controls-and-Systems-Modeling/StructDyn-32656.html Matlab simulators for control and systems modeling]<br />
* [http://www.ae.uiuc.edu/m-selig/ads.html UIUC Airfoil Data Site]<br />
* [http://public.ca.sandia.gov/TNF/abstract.html TNF workshop data archives]<br />
* [http://www.gaussianprocess.org Code for kriging models]<br />
* [http://www.comsol.be/ Comsol multiphysics] package<br />
<br />
=== Predefined functions ===<br />
<br />
* [http://www.cs.cmu.edu/afs/cs/project/jair/pub/volume24/ortizboyer05a-html/node6.html Continuous benchmark problems]<br />
* [http://www2.imm.dtu.dk/~km/Test_ex_forms/test_ex.html Optimization test functions]<br />
* [http://www.geatbx.com/docu/fcnindex-01.html Optimization test functions]<br />
* [http://www.ewh.ieee.org/soc/es/May2001/14/Begin.htm Optimization test functions]<br />
* [http://www.it.lut.fi/ip/evo/functions/functions.html Optimization test functions]<br />
* [http://www.cs.colostate.edu/~genitor/functions.html Optimization test functions]<br />
* [http://www-optima.amp.i.kyoto-u.ac.jp/member/student/hedar/Hedar_files/TestGO_files/Page364.htm Optimization test functions]<br />
* [http://www.it.lut.fi/ip/evo/functions/functions.html Functions with multiple global optima]<br />
* [http://www.mat.univie.ac.at/~neum/glopt/moretest More test set]<br />
* [http://www.mat.univie.ac.at/~neum/software/dixon.tar.gz Dixon-Szegö test set]<br />
* [http://titan.princeton.edu/2010-10-11/TestProblems/ Handbook of Test Problems for local and global optimization]<br />
<br />
=== Optimization ===<br />
<br />
* [http://ab-initio.mit.edu/wiki/index.php/NLopt NLopt is a free/open-source library for nonlinear optimization]<br />
* [http://www.tik.ee.ethz.ch/sop/pisa/ PISA optimization framework]<br />
* [http://www.hvass-labs.org/projects/swarmops/ swarmOps]<br />
* [http://www.icsi.berkeley.edu/~storn/code.html Differential Evolution]<br />
* [http://control.ee.ethz.ch/~joloef/wiki/pmwiki.php YALMIP] High-level optimization problem solver, uses external solvers as backend<br />
* [http://www.mat.univie.ac.at/~neum/glopt/software_g.html Global optimization software]<br />
* [http://www.gerad.ca/NOMAD/Abramson/nomadm.html Mesh-Adaptive Direct Search (MADS) software in Matlab]<br />
* [http://www-rocq.inria.fr/~gilbert/modulopt/modulopt.html Modulopt]: fortran implementations and some matlab<br />
* [http://www.geodise.org/documentation/OptionsMatlab/html/ MatlabOptions] Matlab interface to the design and optimization package (OPTIONS).<br />
<br />
=== Various links ===<br />
<br />
* [http://videolectures.net/Top/Computer_Science/Machine_Learning/ An excellent collection of machine learning lecture videos]<br />
* [http://home.online.no/~pjacklam/matlab/software/util/fullindex.html A collection of useful Matlab scripts]<br />
* [http://home.online.no/~pjacklam/matlab/doc/mtt/ MATLAB array manipulation tips and tricks]<br />
* [http://www.adaptivebox.net/research/bookmark/psocodes_link.html Particle Swarm implementations]<br />
* [http://mloss.org A great selection of open source machine learning software]<br />
* [http://stommel.tamu.edu/~baum/toolboxes.html A list of Matlab toolboxes]<br />
* [http://mdoboard.proboards59.com/ ISSMO-REASON: Research and Engineering Applications in Structural Optimization Network]<br />
* [http://www.nafems.org/about/ NAFEMS]<br />
* [http://www.kat-net.net/ The European Coordinating Action on Key Aerodynamic Technologies]<br />
* [http://www.altairhyperworks.co.uk/Default.aspx Altair Hyperworks]<br />
* [http://www.ifte.de/english/research/index.html IFTE]<br />
* [http://www.optiy.eu/Features.html OptyI]<br />
* [http://www.technet-alliance.com/ Technet Alliance]<br />
<br />
=== Conference links ===<br />
<br />
* [http://www.conferencealerts.com Conference alerts]<br />
* [http://www.conferencealerts.com/engineer.htm Engineering conferences]<br />
* [http://www.conferencealerts.com/ai.htm AI Conference alerts]<br />
* [http://www.ieee.org/web/conferences/search/index.html IEEE conferences]<br />
* [http://www.aiaa.org/content.cfm?pageid=1 AIAA Conferences]<br />
* [http://www.wikicfp.com/ Wiki of conferences]<br />
* [http://users.jyu.fi/~miettine/lista.html#Conferences Kaisa's conferences]<br />
* [http://www.conference-service.com/conferences/neural-networks.html AI Conferences]<br />
* [http://ieee-cis.org/conferences/co_sponsorship_1/ IEEE CIS conferences]<br />
* [http://www.makhfi.com/events.htm Neural Network Events]<br />
* [http://openresearch.org/mw/index.php?title=Upcoming_deadlines&field=Machine+learning Open research AI conferences]<br />
<br />
=== Journal Links ===<br />
<br />
*[http://www.linklings.net/tomacs/charter.html ACM Transactions on Modeling and Computer Simulation]<br />
*[http://www.elsevier.com/wps/find/journaldescription.cws_home/622330/description#description Simulation Modelling Practice and Theory]<br />
*[http://www.elsevier.com/wps/find/journaldescription.cws_home/422911/description#description Advances in Engineering Software]<br />
*[http://journaltool.asme.org/Content/JournalDescriptions.cfm?journalId=12 Journal of Mechanical Design ]<br />
*[http://journaltool.asme.org/Content/JournalDescriptions.cfm?journalId=3&Journal=JCISE Journal of Computing and Information Science in Engineering]<br />
*[http://www.elsevier.com/wps/find/journaldescription.cws_home/975/description#description Engineering Applications of Artificial Intelligence]<br />
*[http://www.elsevier.com/wps/find/journaldescription.cws_home/622240/description#description Advanced Engineering Informatics]<br />
*[http://www.springer.com/computer/information+systems/journal/366 Engineering with Computers]<br />
*[http://www.springer.com/computer/mathematics/journal/521 Neural Computing and Applications]<br />
*[http://www.aiaa.org/content.cfm?pageid=322&lupubid=2 AIAA Journal]<br />
*[http://journals.cambridge.org/action/displayJournal?jid=aie Artificial Intelligence for Engineering Design, Analysis and Manufacturing ]<br />
*[http://jmlr.csail.mit.edu/ Journal of Machine Learning Research]<br />
*[http://www.elsevier.com/wps/find/journaldescription.cws_home/505645/description#description Computer Methods in Applied Mechanics and Engineering]<br />
*[http://www.elsevier.com/wps/find/journaldescription.cws_home/524998/description#description Applied Mathematical Modelling]<br />
*[http://www.siam.org/journals/sisc.php SIAM Journal on Scientific Computing]<br />
*[http://www.iop.org/EJ/journal/-page=scope/0266-5611 Inverse Problems]<br />
*[http://www.tandf.co.uk/journals/titles/17415977.asp Inverse problems in science and engineering]<br />
*[http://ieeexplore.ieee.org/xpl/RecentIssue.jsp?punumber=3468 IEEE Transactions on Systems, Man and Cybernetics, Part A]<br />
*[http://ieeexplore.ieee.org/xpl/RecentIssue.jsp?punumber=3477 IEEE Transactions on Systems, Man and Cybernetics, Part B]<br />
*[http://ieeexplore.ieee.org/xpl/RecentIssue.jsp?punumber=5326 IEEE Transactions on Systems, Man and Cybernetics, Part C]<br />
*[http://www.techscience.com/cmes/aims_scope.html Computer Modeling in Engineering & Sciences]<br />
*[http://www.informaworld.com/smpp/title~db=all~content=t713723652~tab=summary Journal of Experimental & Theoretical Artificial Intelligence]<br />
*[http://www.elsevier.com/locate/jocs Journal of Computational Science]</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=OoDACE:_Citing&diff=5823OoDACE: Citing2012-03-26T10:33:01Z<p>Icouckuy: </p>
<hr />
<div>[[Image:ooDACE.png|250 px|right|ooDACE Toolbox]]<br />
When reporting on results obtained with the ooDACE Toolbox please refer to:<br />
<br />
'''A Surrogate Modeling and Adaptive Sampling Toolbox for Computer Based Design'''<br />
<br>D. Gorissen, K. Crombecq, I. Couckuyt, T. Dhaene, P. Demeester,<br />
<br>Journal of Machine Learning Research,<br />
<br>Vol. 11, pp. 2051−2055, July 2010.<br />
<br>[http://www.jmlr.org/papers/volume11/gorissen10a/gorissen10a.pdf JMRL link]<br />
<br />
<br />
For a list of ooDACE and SUMO related publications see [http://www.sumo.intec.ugent.be/?q=publications The SUMO-lab home page].</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=OoDACE:_License_terms&diff=5822OoDACE: License terms2012-03-26T10:32:50Z<p>Icouckuy: </p>
<hr />
<div>[[Image:osilogo.jpg|90px|right|Open Source Initiative]]<br />
[[Image:Agpl3.png|90px|right|AGPLv3]]<br />
<br />
The ooDACE Toolbox is available under a '''dual license''' model. <br />
<br />
For <font color=red>'''non-commercial'''</font> use, the toolbox is available under the [http://www.fsf.org/licensing/licenses/agpl-3.0.html GNU Affero General Public License version 3] (AGPLv3), an [http://www.opensource.org/ OSI] approved [http://en.wikipedia.org/wiki/Open_source open source] license. <br />
<br />
For use in a <font color=red>'''commercial'''</font> setting, a commercial license must be obtained.<br />
<br />
In addition we require that any reference to the ooDACE Toolbox be accompanied by the [[ooDACE:_Citing|corresponding ooDACE publication]].<br />
<br />
== License terms ==<br />
<br />
[[Image:ooDACE.png|250 px|right|ooDACE Toolbox]]<br />
<br />
For <font color=red>'''non-commercial'''</font> use, this program is free software; you can redistribute it and/or modify it under the terms of the [http://www.fsf.org/licensing/licenses/agpl-3.0.html GNU Affero General Public License version 3] as published by the Free Software Foundation.<br />
<br />
This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Affero General Public License for more details.<br />
<br />
You should have received a copy of the GNU Affero General Public License along with this program; if not, see http://www.gnu.org/licenses or write to the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA, or download the license from the following URL: [http://www.fsf.org/licensing/licenses/agpl-3.0.html http://www.fsf.org/licensing/licenses/agpl-3.0.html]<br />
<br />
In accordance with Section 7(b) of the GNU Affero General Public License, these Appropriate Legal Notices must retain the display of the "'''ooDACE Toolbox'''" text and homepage. In addition, when mentioning the program in written work, reference must be made to the [[ooDACE:_Citing|corresponding ooDACE publication]].<br />
<br />
<br />
You can be released from these requirements by purchasing a <font color=red>'''commercial'''</font> license.<br />
Buying such a license is in most cases mandatory as soon as you develop commercial activities involving the ooDACE Toolbox software. Commercial activities include: consultancy services or using the ooDACE Toolbox in commercial projects (standalone, on a server, through a webservice or other remote access technology).<br />
<br />
For details about a commercial license please [[contact]] us.</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=File:OoDACE.png&diff=5821File:OoDACE.png2012-03-26T10:32:20Z<p>Icouckuy: </p>
<hr />
<div></div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=OoDACE:ooDACE_toolbox&diff=5820OoDACE:ooDACE toolbox2012-03-26T10:32:00Z<p>Icouckuy: </p>
<hr />
<div>== Introduction ==<br />
[[Image:ooDACE.png|250 px|right|ooDACE Toolbox]]<br />
<br />
The ooDACE toolbox is a versatile Matlab toolbox that implements the popular Gaussian Process based kriging surrogate models. Kriging is in particular popular for approximating (and optimizing) deterministic computer experiments. Given a dataset the toolbox automatically fits a kriging surrogate model to it. Afterwards the kriging surrogate can be fully exploited instead of the (probably more expensive) simulation code.<br />
<br />
The toolbox is aimed for solving complex applications (expensive simulation codes, physical experiments, ...) and for researching new kriging extensions and techniques.<br />
<br />
== Download ==<br />
<br />
See: [http://sumo.intec.ugent.be/?q=ooDACE download page]<br />
<br />
== Quick start guide ==<br />
<br />
'''IMPORTANT''': Before the toolbox can be used you have to include the toolbox in Matlab's search path. You can do this manually by running startup, or, if Matlab is started in the root toolbox directory, then startup will be run automatically.<br />
<br />
<source lang="matlab"><br />
startup<br />
</source><br />
<br />
Now the toolbox is ready to be used. The ooDACE toolbox is designed in an object oriented (OO) fashion.<br />
It is strongly recommended to exploit the OO design directly, i.e., use the Kriging and Optimizer matlab classes.<br />
However, for convenience wrapper scripts (dacefit, predictor) are provided that emulate the DACE toolbox interface (see [[#DACE toolbox interface|wrapper scripts]] for more information).<br />
<br />
Assume you want to fit a dataset of n samples in d dimensions.<br />
<br />
<b>samples</b> holds the input parameters n-by-d array (each row is one observation) and <b>values</b> is the corresponding n-by-1 array containing the output values.<br />
<b>lb</b> and <b>ub</b> are 1-by-d arrays defining the lower bounds and upper bounds, respectively, needed to optimize the hyperparameters (<b>theta</b>). In addition, a set of starting values for <b>theta</b> has to be specified (i.e., <b>theta0</b> is also an 1-by-d array)<br />
<br />
As of version 0.2 of the ooDACE toolbox a script is provided, oodacefit, that just takes your dataset (a <b>samples</b> and <b>values</b> matrix) and returns a fitted kriging object, all other parameters (<b>theta0</b>, etc.) are set to some sensible defaults.<br />
<br />
For more flexibility the full example code to fit the dataset is:<br />
<source lang="matlab"><br />
...<br />
% Generate kriging options structure<br />
opts = getDefaultOptions();<br />
opts.hpBounds = [lb ; ub]; % hyperparameter optimization bounds<br />
<br />
% configure the optimization algorithm (only one optimizer is included)<br />
% the Matlab Optimization toolbox is REQUIRED<br />
optimopts.GradObj = 'on';<br />
optimopts.DerivativeCheck = 'off';<br />
optimopts.Diagnostics = 'off';<br />
optimopts.Algorithm = 'active-set';<br />
opts.hpOptimizer = MatlabOptimizer( dim, 1, optimopts );<br />
<br />
% build and fit Kriging object<br />
k = Kriging( opts, theta0, 'regpoly0', @corrgauss );<br />
k = k.fit( samples, values );<br />
<br />
% k represents the approximation and can now be used, e.g.,<br />
[y mse] = k.predict( [1 2] )<br />
...<br />
</source><br />
<br />
See the included demo.m script for more example code on how to use the ooDACE toolbox (including more advanced features such as using blind kriging or how to use regression instead of interpolation). For more information on the classes and their methods please refer to the source files.<br />
<br />
== DACE toolbox interface ==<br />
<br />
The ooDACE toolbox provides two scripts dacefit.m and predictor.m that emulate the behavior of the DACE toolbox ([http://www2.imm.dtu.dk/~hbn/dace/]). Note, that full compatibility between ooDACE and the DACE toolbox is not provided. The scripts merely aim to ease the transition from the DACE toolbox to the ooDACE toolbox.<br />
<br />
Example code:<br />
<source lang="matlab"><br />
krige = dacefit(samples, values, 'regpoly0', 'corrgauss', theta0, lb, ub )<br />
y = predictor([1 2], krige)<br />
</source><br />
<br />
Obviously, a lot less code is used to copy the setup described above. However, less code means less flexibility (e.g., blind kriging and regression kriging are not available using the wrapper scripts). Hence, it is suggested to learn the object oriented interface of ooDACE and use it instead.<br />
<br />
== Contribute ==<br />
<br />
Suggestions on how to improve the ooDACE toolbox are always welcome. For more information please see the [[Feedback | feedback]] page.</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=OoDACE:_Citing&diff=5818OoDACE: Citing2012-03-25T11:06:40Z<p>Icouckuy: </p>
<hr />
<div>[[Image:ooDACE.gif|250 px|right|ooDACE Toolbox]]<br />
When reporting on results obtained with the ooDACE Toolbox please refer to:<br />
<br />
'''A Surrogate Modeling and Adaptive Sampling Toolbox for Computer Based Design'''<br />
<br>D. Gorissen, K. Crombecq, I. Couckuyt, T. Dhaene, P. Demeester,<br />
<br>Journal of Machine Learning Research,<br />
<br>Vol. 11, pp. 2051−2055, July 2010.<br />
<br>[http://www.jmlr.org/papers/volume11/gorissen10a/gorissen10a.pdf JMRL link]<br />
<br />
<br />
For a list of ooDACE and SUMO related publications see [http://www.sumo.intec.ugent.be/?q=publications The SUMO-lab home page].</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=BlindDACE:_Citing&diff=5817BlindDACE: Citing2012-03-25T11:06:26Z<p>Icouckuy: BlindDACE: Citing moved to OoDACE: Citing</p>
<hr />
<div>#REDIRECT [[OoDACE: Citing]]</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=OoDACE:_Citing&diff=5816OoDACE: Citing2012-03-25T11:06:26Z<p>Icouckuy: BlindDACE: Citing moved to OoDACE: Citing</p>
<hr />
<div>[[Image:blindDACE.gif|250 px|right|blindDACE Toolbox]]<br />
When reporting on results obtained with the blindDACE Toolbox please refer to:<br />
<br />
'''A Surrogate Modeling and Adaptive Sampling Toolbox for Computer Based Design'''<br />
<br>D. Gorissen, K. Crombecq, I. Couckuyt, T. Dhaene, P. Demeester,<br />
<br>Journal of Machine Learning Research,<br />
<br>Vol. 11, pp. 2051−2055, July 2010.<br />
<br>[http://www.jmlr.org/papers/volume11/gorissen10a/gorissen10a.pdf JMRL link]<br />
<br />
<br />
For a list of blindDACE and SUMO related publications see [http://www.sumo.intec.ugent.be/?q=publications The SUMO-lab home page].</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=OoDACE:_License_terms&diff=5815OoDACE: License terms2012-03-25T10:57:07Z<p>Icouckuy: </p>
<hr />
<div>[[Image:osilogo.jpg|90px|right|Open Source Initiative]]<br />
[[Image:Agpl3.png|90px|right|AGPLv3]]<br />
<br />
The ooDACE Toolbox is available under a '''dual license''' model. <br />
<br />
For <font color=red>'''non-commercial'''</font> use, the toolbox is available under the [http://www.fsf.org/licensing/licenses/agpl-3.0.html GNU Affero General Public License version 3] (AGPLv3), an [http://www.opensource.org/ OSI] approved [http://en.wikipedia.org/wiki/Open_source open source] license. <br />
<br />
For use in a <font color=red>'''commercial'''</font> setting, a commercial license must be obtained.<br />
<br />
In addition we require that any reference to the ooDACE Toolbox be accompanied by the [[ooDACE:_Citing|corresponding ooDACE publication]].<br />
<br />
== License terms ==<br />
<br />
[[Image:ooDACE.gif|250 px|right|ooDACE Toolbox]]<br />
<br />
For <font color=red>'''non-commercial'''</font> use, this program is free software; you can redistribute it and/or modify it under the terms of the [http://www.fsf.org/licensing/licenses/agpl-3.0.html GNU Affero General Public License version 3] as published by the Free Software Foundation.<br />
<br />
This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Affero General Public License for more details.<br />
<br />
You should have received a copy of the GNU Affero General Public License along with this program; if not, see http://www.gnu.org/licenses or write to the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA, or download the license from the following URL: [http://www.fsf.org/licensing/licenses/agpl-3.0.html http://www.fsf.org/licensing/licenses/agpl-3.0.html]<br />
<br />
In accordance with Section 7(b) of the GNU Affero General Public License, these Appropriate Legal Notices must retain the display of the "'''ooDACE Toolbox'''" text and homepage. In addition, when mentioning the program in written work, reference must be made to the [[ooDACE:_Citing|corresponding ooDACE publication]].<br />
<br />
<br />
You can be released from these requirements by purchasing a <font color=red>'''commercial'''</font> license.<br />
Buying such a license is in most cases mandatory as soon as you develop commercial activities involving the ooDACE Toolbox software. Commercial activities include: consultancy services or using the ooDACE Toolbox in commercial projects (standalone, on a server, through a webservice or other remote access technology).<br />
<br />
For details about a commercial license please [[contact]] us.</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=BlindDACE:_License_terms&diff=5814BlindDACE: License terms2012-03-25T10:56:28Z<p>Icouckuy: BlindDACE: License terms moved to OoDACE: License terms: renamed toolbox</p>
<hr />
<div>#REDIRECT [[OoDACE: License terms]]</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=OoDACE:_License_terms&diff=5813OoDACE: License terms2012-03-25T10:56:28Z<p>Icouckuy: BlindDACE: License terms moved to OoDACE: License terms: renamed toolbox</p>
<hr />
<div>[[Image:osilogo.jpg|90px|right|Open Source Initiative]]<br />
[[Image:Agpl3.png|90px|right|AGPLv3]]<br />
<br />
The blindDACE Toolbox is available under a '''dual license''' model. <br />
<br />
For <font color=red>'''non-commercial'''</font> use, the toolbox is available under the [http://www.fsf.org/licensing/licenses/agpl-3.0.html GNU Affero General Public License version 3] (AGPLv3), an [http://www.opensource.org/ OSI] approved [http://en.wikipedia.org/wiki/Open_source open source] license. <br />
<br />
For use in a <font color=red>'''commercial'''</font> setting, a commercial license must be obtained.<br />
<br />
In addition we require that any reference to the blindDACE Toolbox be accompanied by the [[BlindDACE:_Citing|corresponding blindDACE publication]].<br />
<br />
== License terms ==<br />
<br />
[[Image:blindDACE.gif|250 px|right|blindDACE Toolbox]]<br />
<br />
For <font color=red>'''non-commercial'''</font> use, this program is free software; you can redistribute it and/or modify it under the terms of the [http://www.fsf.org/licensing/licenses/agpl-3.0.html GNU Affero General Public License version 3] as published by the Free Software Foundation.<br />
<br />
This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Affero General Public License for more details.<br />
<br />
You should have received a copy of the GNU Affero General Public License along with this program; if not, see http://www.gnu.org/licenses or write to the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA, or download the license from the following URL: [http://www.fsf.org/licensing/licenses/agpl-3.0.html http://www.fsf.org/licensing/licenses/agpl-3.0.html]<br />
<br />
In accordance with Section 7(b) of the GNU Affero General Public License, these Appropriate Legal Notices must retain the display of the "'''blindDACE Toolbox'''" text and homepage. In addition, when mentioning the program in written work, reference must be made to the [[BlindDACE:_Citing|corresponding blindDACE publication]].<br />
<br />
<br />
You can be released from these requirements by purchasing a <font color=red>'''commercial'''</font> license.<br />
Buying such a license is in most cases mandatory as soon as you develop commercial activities involving the blindDACE Toolbox software. Commercial activities include: consultancy services or using the blindDACE Toolbox in commercial projects (standalone, on a server, through a webservice or other remote access technology).<br />
<br />
For details about a commercial license please [[contact]] us.</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=OoDACE:ooDACE_toolbox&diff=5812OoDACE:ooDACE toolbox2012-03-25T10:55:46Z<p>Icouckuy: </p>
<hr />
<div>== Introduction ==<br />
[[Image:ooDACE.gif|250 px|right|ooDACE Toolbox]]<br />
<br />
The ooDACE toolbox is a versatile Matlab toolbox that implements the popular Gaussian Process based kriging surrogate models. Kriging is in particular popular for approximating (and optimizing) deterministic computer experiments. Given a dataset the toolbox automatically fits a kriging surrogate model to it. Afterwards the kriging surrogate can be fully exploited instead of the (probably more expensive) simulation code.<br />
<br />
The toolbox is aimed for solving complex applications (expensive simulation codes, physical experiments, ...) and for researching new kriging extensions and techniques.<br />
<br />
== Download ==<br />
<br />
See: [http://sumo.intec.ugent.be/?q=ooDACE download page]<br />
<br />
== Quick start guide ==<br />
<br />
'''IMPORTANT''': Before the toolbox can be used you have to include the toolbox in Matlab's search path. You can do this manually by running startup, or, if Matlab is started in the root toolbox directory, then startup will be run automatically.<br />
<br />
<source lang="matlab"><br />
startup<br />
</source><br />
<br />
Now the toolbox is ready to be used. The ooDACE toolbox is designed in an object oriented (OO) fashion.<br />
It is strongly recommended to exploit the OO design directly, i.e., use the Kriging and Optimizer matlab classes.<br />
However, for convenience wrapper scripts (dacefit, predictor) are provided that emulate the DACE toolbox interface (see [[#DACE toolbox interface|wrapper scripts]] for more information).<br />
<br />
Assume you want to fit a dataset of n samples in d dimensions.<br />
<br />
<b>samples</b> holds the input parameters n-by-d array (each row is one observation) and <b>values</b> is the corresponding n-by-1 array containing the output values.<br />
<b>lb</b> and <b>ub</b> are 1-by-d arrays defining the lower bounds and upper bounds, respectively, needed to optimize the hyperparameters (<b>theta</b>). In addition, a set of starting values for <b>theta</b> has to be specified (i.e., <b>theta0</b> is also an 1-by-d array)<br />
<br />
As of version 0.2 of the ooDACE toolbox a script is provided, oodacefit, that just takes your dataset (a <b>samples</b> and <b>values</b> matrix) and returns a fitted kriging object, all other parameters (<b>theta0</b>, etc.) are set to some sensible defaults.<br />
<br />
For more flexibility the full example code to fit the dataset is:<br />
<source lang="matlab"><br />
...<br />
% Generate kriging options structure<br />
opts = getDefaultOptions();<br />
opts.hpBounds = [lb ; ub]; % hyperparameter optimization bounds<br />
<br />
% configure the optimization algorithm (only one optimizer is included)<br />
% the Matlab Optimization toolbox is REQUIRED<br />
optimopts.GradObj = 'on';<br />
optimopts.DerivativeCheck = 'off';<br />
optimopts.Diagnostics = 'off';<br />
optimopts.Algorithm = 'active-set';<br />
opts.hpOptimizer = MatlabOptimizer( dim, 1, optimopts );<br />
<br />
% build and fit Kriging object<br />
k = Kriging( opts, theta0, 'regpoly0', @corrgauss );<br />
k = k.fit( samples, values );<br />
<br />
% k represents the approximation and can now be used, e.g.,<br />
[y mse] = k.predict( [1 2] )<br />
...<br />
</source><br />
<br />
See the included demo.m script for more example code on how to use the ooDACE toolbox (including more advanced features such as using blind kriging or how to use regression instead of interpolation). For more information on the classes and their methods please refer to the source files.<br />
<br />
== DACE toolbox interface ==<br />
<br />
The ooDACE toolbox provides two scripts dacefit.m and predictor.m that emulate the behavior of the DACE toolbox ([http://www2.imm.dtu.dk/~hbn/dace/]). Note, that full compatibility between ooDACE and the DACE toolbox is not provided. The scripts merely aim to ease the transition from the DACE toolbox to the ooDACE toolbox.<br />
<br />
Example code:<br />
<source lang="matlab"><br />
krige = dacefit(samples, values, 'regpoly0', 'corrgauss', theta0, lb, ub )<br />
y = predictor([1 2], krige)<br />
</source><br />
<br />
Obviously, a lot less code is used to copy the setup described above. However, less code means less flexibility (e.g., blind kriging and regression kriging are not available using the wrapper scripts). Hence, it is suggested to learn the object oriented interface of ooDACE and use it instead.<br />
<br />
== Contribute ==<br />
<br />
Suggestions on how to improve the ooDACE toolbox are always welcome. For more information please see the [[Feedback | feedback]] page.</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=BlindDACE:BlindDACE_toolbox&diff=5811BlindDACE:BlindDACE toolbox2012-03-25T10:55:07Z<p>Icouckuy: BlindDACE:BlindDACE toolbox moved to OoDACE:ooDACE toolbox: renamed toolbox</p>
<hr />
<div>#REDIRECT [[OoDACE:ooDACE toolbox]]</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=OoDACE:ooDACE_toolbox&diff=5810OoDACE:ooDACE toolbox2012-03-25T10:55:07Z<p>Icouckuy: BlindDACE:BlindDACE toolbox moved to OoDACE:ooDACE toolbox: renamed toolbox</p>
<hr />
<div>== Introduction ==<br />
[[Image:blindDACE.gif|250 px|right|blindDACE Toolbox]]<br />
<br />
The blindDACE toolbox is a versatile Matlab toolbox that implements the popular Gaussian Process based kriging surrogate models. Kriging is in particular popular for approximating (and optimizing) deterministic computer experiments. Given a dataset the toolbox automatically fits a kriging surrogate model to it. Afterwards the kriging surrogate can be fully exploited instead of the (probably more expensive) simulation code.<br />
<br />
The toolbox is aimed for solving complex applications (expensive simulation codes, physical experiments, ...) and for researching new kriging extensions and techniques.<br />
<br />
== Download ==<br />
<br />
See: [http://sumo.intec.ugent.be/?q=blindDACE download page]<br />
<br />
== Quick start guide ==<br />
<br />
'''IMPORTANT''': Before the toolbox can be used you have to include the toolbox in Matlab's search path. You can do this manually by running startup, or, if Matlab is started in the root toolbox directory, then startup will be run automatically.<br />
<br />
<source lang="matlab"><br />
startup<br />
</source><br />
<br />
Now the toolbox is ready to be used. The blindDACE toolbox is designed in an object oriented (OO) fashion.<br />
It is strongly recommended to exploit the OO design directly, i.e., use the Kriging and Optimizer matlab classes.<br />
However, for convenience wrapper scripts (dacefit, predictor) are provided that emulate the DACE toolbox interface (see [[#DACE toolbox interface|wrapper scripts]] for more information).<br />
<br />
Assume you want to fit a dataset of n samples in d dimensions.<br />
<br />
<b>samples</b> holds the input parameters n-by-d array (each row is one observation) and <b>values</b> is the corresponding n-by-1 array containing the output values.<br />
<b>lb</b> and <b>ub</b> are 1-by-d arrays defining the lower bounds and upper bounds, respectively, needed to optimize the hyperparameters (<b>theta</b>). In addition, a set of starting values for <b>theta</b> has to be specified (i.e., <b>theta0</b> is also an 1-by-d array)<br />
<br />
As of version 0.2 of the blindDACE toolbox a script is provided, blinddacefit, that just takes your dataset (a <b>samples</b> and <b>values</b> matrix) and returns a fitted kriging object, all other parameters (<b>theta0</b>, etc.) are set to some sensible defaults.<br />
<br />
For more flexibility the full example code to fit the dataset is:<br />
<source lang="matlab"><br />
...<br />
% Generate kriging options structure<br />
opts = getDefaultOptions();<br />
opts.hpBounds = [lb ; ub]; % hyperparameter optimization bounds<br />
<br />
% configure the optimization algorithm (only one optimizer is included)<br />
% the Matlab Optimization toolbox is REQUIRED<br />
optimopts.GradObj = 'on';<br />
optimopts.DerivativeCheck = 'off';<br />
optimopts.Diagnostics = 'off';<br />
optimopts.Algorithm = 'active-set';<br />
opts.hpOptimizer = MatlabOptimizer( dim, 1, optimopts );<br />
<br />
% build and fit Kriging object<br />
k = Kriging( opts, theta0, 'regpoly0', @corrgauss );<br />
k = k.fit( samples, values );<br />
<br />
% k represents the approximation and can now be used, e.g.,<br />
[y mse] = k.predict( [1 2] )<br />
...<br />
</source><br />
<br />
See the included demo.m script for more example code on how to use the blindDACE toolbox (including more advanced features such as using blind kriging or how to use regression instead of interpolation). For more information on the classes and their methods please refer to the source files.<br />
<br />
== DACE toolbox interface ==<br />
<br />
The blindDACE toolbox provides two scripts dacefit.m and predictor.m that emulate the behavior of the DACE toolbox ([http://www2.imm.dtu.dk/~hbn/dace/]). Note, that full compatibility between blindDACE and the DACE toolbox is not provided. The scripts merely aim to ease the transition from the DACE toolbox to the blindDACE toolbox.<br />
<br />
Example code:<br />
<source lang="matlab"><br />
krige = dacefit(samples, values, 'regpoly0', 'corrgauss', theta0, lb, ub )<br />
y = predictor([1 2], krige)<br />
</source><br />
<br />
Obviously, a lot less code is used to copy the setup described above. However, less code means less flexibility (e.g., blind kriging and regression kriging are not available using the wrapper scripts). Hence, it is suggested to learn the object oriented interface of blindDACE and use it instead.<br />
<br />
== Contribute ==<br />
<br />
Suggestions on how to improve the blindDACE toolbox are always welcome. For more information please see the [[Feedback | feedback]] page.</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=OoDACE:ooDACE_toolbox&diff=5225OoDACE:ooDACE toolbox2010-10-08T08:59:52Z<p>Icouckuy: /* Quick start guide */</p>
<hr />
<div>== Introduction ==<br />
<br />
The blindDACE toolbox is a versatile Matlab toolbox that implements the popular Gaussian Process based kriging surrogate models. Kriging is in particular popular for approximating (and optimizing) deterministic computer experiments. Given a dataset the toolbox automatically fits a kriging surrogate model to it. Afterwards the kriging surrogate can be fully exploited instead of the (probably more expensive) simulation code.<br />
<br />
The toolbox is aimed for solving complex applications (expensive simulation codes, physical experiments, ...) and for researching new kriging extensions and techniques.<br />
<br />
== Download ==<br />
<br />
See: [http://sumo.intec.ugent.be/?q=blindDACE download page]<br />
<br />
== Quick start guide ==<br />
<br />
'''IMPORTANT''': Before the toolbox can be used you have to include the toolbox in Matlab's search path. You can do this manually by running startup, or, if Matlab is started in the root toolbox directory, then startup will be run automatically.<br />
<br />
<source lang="matlab"><br />
startup<br />
</source><br />
<br />
Now the toolbox is ready to be used. The blindDACE toolbox is designed in an object oriented (OO) fashion.<br />
It is strongly recommended to exploit the OO design directly, i.e., use the Kriging and Optimizer matlab classes.<br />
However, for convenience wrapper scripts (dacefit, predictor) are provided that emulate the DACE toolbox interface (see [[#DACE toolbox interface|wrapper scripts]] for more information).<br />
<br />
Assume you want to fit a dataset of n samples in d dimensions.<br />
<br />
<b>samples</b> holds the input parameters n-by-d array (each row is one observation) and <b>values</b> is the corresponding n-by-1 array containing the output values.<br />
<b>lb</b> and <b>ub</b> are 1-by-d arrays defining the lower bounds and upper bounds, respectively, needed to optimize the hyperparameters (<b>theta</b>). In addition, a set of starting values for <b>theta</b> has to be specified (i.e., <b>theta0</b> is also an 1-by-d array)<br />
<br />
As of version 0.2 of the blindDACE toolbox a script is provided, blinddacefit, that just takes your dataset (a <b>samples</b> and <b>values</b> matrix) and returns a fitted kriging object, all other parameters (<b>theta0</b>, etc.) are set to some sensible defaults.<br />
<br />
For more flexibility the full example code to fit the dataset is:<br />
<source lang="matlab"><br />
...<br />
% Generate kriging options structure<br />
opts = getDefaultOptions();<br />
opts.hpBounds = [lb ; ub]; % hyperparameter optimization bounds<br />
<br />
% configure the optimization algorithm (only one optimizer is included)<br />
% the Matlab Optimization toolbox is REQUIRED<br />
optimopts.GradObj = 'on';<br />
optimopts.DerivativeCheck = 'off';<br />
optimopts.Diagnostics = 'off';<br />
optimopts.Algorithm = 'active-set';<br />
opts.hpOptimizer = MatlabOptimizer( dim, 1, optimopts );<br />
<br />
% build and fit Kriging object<br />
k = Kriging( opts, theta0, 'regpoly0', @corrgauss );<br />
k = k.fit( samples, values );<br />
<br />
% k represents the approximation and can now be used, e.g.,<br />
[y mse] = k.predict( [1 2] )<br />
...<br />
</source><br />
<br />
See the included demo.m script for more example code on how to use the blindDACE toolbox (including more advanced features such as using blind kriging or how to use regression instead of interpolation). For more information on the classes and their methods please refer to the source files.<br />
<br />
== DACE toolbox interface ==<br />
<br />
The blindDACE toolbox provides two scripts dacefit.m and predictor.m that emulate the behavior of the DACE toolbox ([http://www2.imm.dtu.dk/~hbn/dace/]). Note, that full compatibility between blindDACE and the DACE toolbox is not provided. The scripts merely aim to ease the transition from the DACE toolbox to the blindDACE toolbox.<br />
<br />
Example code:<br />
<source lang="matlab"><br />
krige = dacefit(samples, values, 'regpoly0', 'corrgauss', theta0, lb, ub )<br />
y = predictor([1 2], krige)<br />
</source><br />
<br />
Obviously, a lot less code is used to copy the setup described above. However, less code means less flexibility (e.g., blind kriging and regression kriging are not available using the wrapper scripts). Hence, it is suggested to learn the object oriented interface of blindDACE and use it instead.<br />
<br />
== Contribute ==<br />
<br />
Suggestions on how to improve the blindDACE toolbox are always welcome. For more information please see the [[Feedback | feedback]] page.</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=OoDACE:ooDACE_toolbox&diff=5224OoDACE:ooDACE toolbox2010-10-08T08:56:43Z<p>Icouckuy: /* Quick start guide */</p>
<hr />
<div>== Introduction ==<br />
<br />
The blindDACE toolbox is a versatile Matlab toolbox that implements the popular Gaussian Process based kriging surrogate models. Kriging is in particular popular for approximating (and optimizing) deterministic computer experiments. Given a dataset the toolbox automatically fits a kriging surrogate model to it. Afterwards the kriging surrogate can be fully exploited instead of the (probably more expensive) simulation code.<br />
<br />
The toolbox is aimed for solving complex applications (expensive simulation codes, physical experiments, ...) and for researching new kriging extensions and techniques.<br />
<br />
== Download ==<br />
<br />
See: [http://sumo.intec.ugent.be/?q=blindDACE download page]<br />
<br />
== Quick start guide ==<br />
<br />
'''IMPORTANT''': Before the toolbox can be used you have to include the toolbox in Matlab's search path. You can do this manually by running startup, or, if Matlab is started in the root toolbox directory, then startup will be run automatically.<br />
<br />
<source lang="matlab"><br />
startup<br />
</source><br />
<br />
Now the toolbox is ready to be used. The blindDACE toolbox is designed in an object oriented (OO) fashion.<br />
It is strongly recommended to exploit the OO design directly, i.e., use the Kriging and Optimizer matlab classes.<br />
However, for convenience wrapper scripts (dacefit, predictor) are provided that emulate the DACE toolbox interface (see [[#DACE toolbox interface|wrapper scripts]] for more information).<br />
<br />
Assume you want to fit a dataset of n samples in d dimensions.<br />
<br />
<b>samples</b> holds the input parameters n-by-d array (each row is one observation) and <b>values</b> is the corresponding n-by-1 array containing the output values.<br />
<b>lb</b> and <b>ub</b> are 1-by-d arrays defining the lower bounds and upper bounds, respectively, needed to optimize the hyperparameters. In addition, a set of starting values has to be specified (e.g., <b>theta0</b> is also an 1-by-d array)<br />
<br />
As of version 0.2 of the blindDACE toolbox a script is provided, blinddacefit, that just takes your dataset (a <b>samples</b> and <b>values</b> matrix) and returns a fitted kriging object, all other parameters (<b>theta0</b>, etc.) are set to some sensible defaults.<br />
<br />
For more flexibility the full example code to fit the dataset is:<br />
<source lang="matlab"><br />
...<br />
% Generate kriging options structure<br />
opts = getDefaultOptions();<br />
opts.hpBounds = [lb ; ub]; % hyperparameter optimization bounds<br />
<br />
% configure the optimization algorithm (only one optimizer is included)<br />
% the Matlab Optimization toolbox is REQUIRED<br />
optimopts.GradObj = 'on';<br />
optimopts.DerivativeCheck = 'off';<br />
optimopts.Diagnostics = 'off';<br />
optimopts.Algorithm = 'active-set';<br />
opts.hpOptimizer = MatlabOptimizer( dim, 1, optimopts );<br />
<br />
% build and fit Kriging object<br />
k = Kriging( opts, theta0, 'regpoly0', @corrgauss );<br />
k = k.fit( samples, values );<br />
<br />
% k represents the approximation and can now be used, e.g.,<br />
[y mse] = k.predict( [1 2] )<br />
...<br />
</source><br />
<br />
See the included demo.m script for more example code on how to use the blindDACE toolbox (including more advanced features such as using blind kriging or how to use regression instead of interpolation). For more information on the classes and their methods please refer to the source files.<br />
<br />
== DACE toolbox interface ==<br />
<br />
The blindDACE toolbox provides two scripts dacefit.m and predictor.m that emulate the behavior of the DACE toolbox ([http://www2.imm.dtu.dk/~hbn/dace/]). Note, that full compatibility between blindDACE and the DACE toolbox is not provided. The scripts merely aim to ease the transition from the DACE toolbox to the blindDACE toolbox.<br />
<br />
Example code:<br />
<source lang="matlab"><br />
krige = dacefit(samples, values, 'regpoly0', 'corrgauss', theta0, lb, ub )<br />
y = predictor([1 2], krige)<br />
</source><br />
<br />
Obviously, a lot less code is used to copy the setup described above. However, less code means less flexibility (e.g., blind kriging and regression kriging are not available using the wrapper scripts). Hence, it is suggested to learn the object oriented interface of blindDACE and use it instead.<br />
<br />
== Contribute ==<br />
<br />
Suggestions on how to improve the blindDACE toolbox are always welcome. For more information please see the [[Feedback | feedback]] page.</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=Useful_Links&diff=5113Useful Links2010-03-25T12:44:04Z<p>Icouckuy: /* Optimization */</p>
<hr />
<div>=== Related publications ===<br />
<br />
See the [[Related publications]] page.<br />
<br />
=== Related projects ===<br />
<br />
A list of projects with similar ideas and scope.<br />
* [http://www.cs.rtu.lv/jekabsons/regression.html Matlab regression software]<br />
* [http://www.nutechsolutions.com/prod_cv_analytics.asp ClearVu Analytics]<br />
* [http://fchegury.googlepages.com/surrogatestoolbox Surrogates Toolbox]: Another matlab surrogate modeling toolbox<br />
* [http://www.evolved-analytics.com Evolved-Analytics]: DataModeler package.<br />
* [http://www.muneda.com/ MunEDA]: MunEDA provides leading EDA software technology for analysis, modelling and optimization of yield and performance of analog, mixed-signal and digital designs.<br />
* [http://www.lsoptsupport.com/faqs/setting-parameters-for-metamodel-based-optimization-strategies LS-OPT: Functionality similar to the SUMO-Toolbox in LS-OPT]<br />
* [http://mucm.aston.ac.uk/MUCM/MUCMToolkit/index.php?page=MetaHomePage.html MUCH Toolkit]: A toolbox for sensitivity analysis using surrogate models<br />
* [http://home.mit.bme.hu/~kollar/topics/fdident.html FDIDENT: frequency domain identification toolbox]<br />
* [http://www.dmoz.org/Science/Math/Statistics/Software/Regression_and_Curve_Fitting/: Regression software on Open Directory]<br />
* [http://www.infiniscale.com/ Infiniscale]: TechModeler, TechAnalyzer: automatic model generation tools for EM applications<br />
* [http://www.vrand.com/visualDOC.html visualDOC]: VisualDOC, a design optimization tool<br />
* [http://www.salford-systems.com/mars.php MARS]: A spline based modeling tool<br />
* [http://www.phoenix-int.com/products/modelcenter.php Modelcenter]: A datamining and modeling tool<br />
* [http://www.engineous.com/product_iSIGHT.htm iSIGHT]: A datamining and modeling tool<br />
* [http://www.tmpinc.com/datascape_overview.html Datascape]: A datamining and modeling tool<br />
* [http://www.friendship-systems.com/ Friendship systems]: Tools for modeling ship hulls parametrically and performing the modeling calculations (Equilibrium tool)<br />
* [http://www.csse.monash.edu.au/%7Edavida/nimrod/nimrodg.htm Nimrod/G]: Execute parameter sweeps on the grid<br />
* [http://www-sop.inria.fr/oasis/ProActive/ ProActive]: Java grid library <br />
* [http://www.csse.monash.edu.au/%7Edavida/nimrod/nimrodo.htm Nimrod/O]: Grid-enabled optimization toolkit <br />
* [http://www.ece.northwestern.edu/OTC/ NEOS]: Distributed optimization toolkit <br />
* [http://www.nas.nasa.gov/SC2000/ARC/ilab.html iLab]: automated parameter study toolkit <br />
* [http://software.sci.utah.edu/scirun.html SciRun]: Problem Solving Environment (PSE), for simulation, modeling, and visualization of scientific problems. <br />
* [http://www.esteco.it/ ModeFRONTIER]: environment dedicated to the set up of design assessment chains and efficient investigation of the design space. <br />
* [http://www.gridbus.org/ Gridbus]: metascheduler for the grid <br />
* [http://icl.cs.utk.edu/netsolve/overview/ Gridsolve/Netsolve]: grid enabled scientific computation toolbox <br />
* [http://www.dtreg.com/ DTREG]: a powerful statistical analysis program that generates classification and regression trees and Support Vector Machine models that can be used to predict parameter values. <br />
* [http://www.soton.ac.uk/%7Epbn/MDO/ MDO]: A collection of MDO links <br />
* [http://www.research.att.com/~njas/gosset/index.html GOSSET]: A general purpose program for designing experiments <br />
* [http://www.cs.sandia.gov/DAKOTA/ DAKOTA]: Design Analysis Kit for Optimization and Terascale Applications<br />
* [http://www.wesc.ac.uk/projectsite/dipso/index.html DIPSO]: Wide-Area Distributed Problem Solving (DIPSO <br />
* [http://www.geodise.org GEODISE]: Grid Enabled Optimisation and Design Search for Engineering<br />
* [http://czms.mit.edu/poseidon/new1/ Poseidon]: A distributed information system for ocean processes.<br />
* [http://www.fast.u-psud.fr/ezyfit/ EZfit] : Free curve fitting toolbox for matlab<br />
* [http://www.ians.uni-stuttgart.de/spinterp/about.html SGIT] : A Sparse grid interpolation toolbox<br />
* [http://www.csie.ntu.edu.tw/~yien/quickrbf/quickstart.php QuickRBF] : an RBF fitting library (native)<br />
* [http://www.farfieldtechnology.com/products/toolbox/ FastRBF] : another RBF fitting library (matlab)<br />
* [http://www.neuromat.com/models.html Neuromat Predictor] : neural networks fitting library<br />
* [http://simlab.jrc.ec.europa.eu/ Sensitivity Analysis library]<br />
* Gaussian Process Matlab code<br />
** [http://www.gaussianprocess.org/gpml/code/matlab/doc/ Code] based on Rasmussen's book<br />
** [http://www.kyb.mpg.de/publication.html?publ=2689 Sparse Gaussian Processes]<br />
** [http://www.cs.man.ac.uk/~neill/gp/ Gaussian Process Software]<br />
** [http://www.ios.htwg-konstanz.de/joomla_mof/index.php?option=com_content&view=article&id=48:polyreg-polynomial-gaussian-process-regression&catid=36:code&Itemid=81 GP] using polynomial covariance functions<br />
<br />
=== Related labs ===<br />
<br />
A list of some of the labs/researchers with similar ideas and scope.<br />
<br />
* [http://aerospace.engin.umich.edu/index.html Department of Aerospace Engineering at the University of Michigan]<br />
* [http://www.mae.ufl.edu/~mdo/research.html The Structural and Multidisciplinary Optimization Group at the University of Florida]<br />
* [http://web.engr.oregonstate.edu/~tgd/ School of Electrical Engineering and Computer Science, Oregon State U]<br />
* [http://www.soton.ac.uk/~cedc/ Computational Engineering and Design Center]<br />
* [http://edog.mne.psu.edu/research.html Engineering Design & Optimization Group (Penn state)]<br />
* [http://www.nd.edu/~sgano/research.html Aerospace and Mechanical Engineering] [http://www.gano.name/shawn/ Homepage]<br />
* [http://shyylab.engin.umich.edu/research/design-optimization Computational Thermo-Fluids Group]<br />
* [http://www.ensc.sfu.ca/~gwa5/index.htm Product Design and Optimization Laboratory (PDOL)]<br />
* [http://www.cerfacs.fr/4-25708-Home.php Computational Fluid Dynamics (CFD) group at CERFACS]<br />
* [http://webuser.uni-weimar.de/~roos1/ Dirk Roos]<br />
* [http://www.cerfacs.fr/~duchaine/HTML/research.htm Florent Duchaine]<br />
* http://www.nlr.nl/ National aerospace lab]<br />
<br />
=== Data sets - Simulation code ===<br />
<br />
A list of publicaly available datasets and simulation codes, useful for testing.<br />
<br />
* [http://matlabdb.mathematik.uni-stuttgart.de/files.jsp?MC_ID=1&SC_ID=2 Matlab scientific computing database] : Nicly documented Matlab simulation code examples<br />
* [http://www.mat.univie.ac.at/~neum/stat.html Statistics links] : A nice collection of data fitting and analysis codes<br />
* [http://en.wikipedia.org/wiki/Classic_data_sets Wikipedia Classic Datasets]<br />
* [http://www.uni-koeln.de/themen/Statistik/data/rousseeuw/ The ROUSSEEUW datasets]<br />
* [ftp://ftp.sas.com/pub/neural/dojo/dojo.html Donoho-Johnstone Benchmarks] <br />
* [http://people.scs.fsu.edu/~burkardt/datasets/datasets.html datasets]<br />
* [http://www.cise.ufl.edu/~mpf/sch/ Schrödinger wave simulations]<br />
* [http://www-syscom.univ-mlv.fr/~vignat/Signal/Space/index.html Spice like simulator for matlab]<br />
* [http://www.itl.nist.gov/div898/strd/general/dataarchive.html Nist dataset archive]<br />
* [http://homes.esat.kuleuven.be/~smc/daisy/daisydata.html Daisy datasets]<br />
* [http://lib.stat.cmu.edu/datasets/ Statlib dataset archive]<br />
* [http://www.iau.dtu.dk/nnbook/systems.html Datasets from the book "Neural networks for the modeling and control of dynamic systems.]<br />
* [http://www.gpc.de/eposes.html A simulation environment for production and transport, logistics and automation systems]<br />
* [http://phy.asu.edu/shumway/codes.html Nanostructure Simulation and Modeling Programs]<br />
* [http://pedsim.silmaril.org/ A Modular, Distributed Pedestrian Crowd Simulation System]<br />
* [http://g95.sourceforge.net/g95_status.html Fortran simulation codes]<br />
* [http://www.genie.ac.uk/ Grid ENabled Integrated Earth system model]<br />
* [http://gcmd.nasa.gov/KeywordSearch/Home.do?Portal=GCMD&MetadataType=0 Nasa datasets]<br />
* [http://funapp.cs.bilkent.edu.tr/DataSets/ Function approximation repository]<br />
* [http://www.google.com/Top/Computers/Artificial_Intelligence/Machine_Learning/Datasets/ Google directory datasets]<br />
* [http://www.idsia.ch/~andrea/simtools.html A Collection of Modeling and Simulation Resources on the Internet]<br />
* [http://www.grc.nasa.gov/WWW/K-12/freesoftware_page.htm Free simulation software from Nasa]<br />
* [http://opensees.berkeley.edu/index.php Earthquake simulation]<br />
* [http://www.pdl.cmu.edu/DiskSim/ Harddisk simulator]<br />
* [http://spib.rice.edu/spib/mtn_top.html Mountain top radar data]<br />
* [http://www.statsci.org/datasets.html Statsci dataset repository]<br />
* [http://statwww.epfl.ch/davison/BMA/Data4BMA/ dataset repository]<br />
* [http://astrostatistics.psu.edu/datasets/ dataset repository]<br />
* [http://www.maths.uq.edu.au/CEToolBox/ problems from the Cross Entropy toolbox]<br />
* [http://www.cs.waikato.ac.nz/~ml/weka/index_datasets.html Weka datasets]<br />
* [http://www.ailab.si/orange/datasets.asp?Inst=on&Atts=on&Class=on&Values=on&Description=on&sort=Data+Set dataset repository]<br />
* [http://www.itee.uq.edu.au/%7Emarcusg/msg.html Max Set of Gaussians Landscape Generator]<br />
* [http://www.mathworks.fr/matlabcentral/fileexchange/loadAuthor.do?objectId=364966&objectType=author Collection of useful Matlab scripts by Thomas Abrahamsson ]<br />
* [http://www.mathworks.com/matlabcentral/fileexchange/loadFile.do?objectId=10731&objectType=FILE# A Matlab toolbox for Linear Structural Dynamics Analysis]<br />
* [http://webscripts.softpedia.com/script/Scientific-Engineering-Ruby/Controls-and-Systems-Modeling/StructDyn-32656.html Matlab simulators for control and systems modeling]<br />
* [http://www.ae.uiuc.edu/m-selig/ads.html UIUC Airfoil Data Site]<br />
* [http://citeseer.ist.psu.edu/112408.html Proben1 benchmark datasets]<br />
* [http://public.ca.sandia.gov/TNF/abstract.html TNF workshop data archives]<br />
* [http://www.gaussianprocess.org Code for kriging models]<br />
* [http://www.comsol.be/ Comsol multiphysics] package<br />
<br />
=== Predefined functions ===<br />
<br />
* [http://www.cs.cmu.edu/afs/cs/project/jair/pub/volume24/ortizboyer05a-html/node6.html Continuous benchmark problems]<br />
* [http://www2.imm.dtu.dk/~km/GlobOpt/testex/ Optimization test functions]<br />
* [http://www.geatbx.com/docu/fcnindex-01.html Optimization test functions]<br />
* [http://www.ewh.ieee.org/soc/es/May2001/14/Begin.htm Optimization test functions]<br />
* [http://www.it.lut.fi/ip/evo/functions/functions.html Optimization test functions]<br />
* [http://www.cs.colostate.edu/~genitor/functions.html Optimization test functions]<br />
* [http://www-optima.amp.i.kyoto-u.ac.jp/member/student/hedar/Hedar_files/TestGO_files/Page364.htm Optimization test functions]<br />
* [http://ntu-cg.ntu.edu.sg/ysong/journal/GLSurrogate.pdf Surrogate modeling test functions]<br />
* [http://www.it.lut.fi/ip/evo/functions/functions.html Functions with multiple global optima]<br />
* [http://www.mat.univie.ac.at/~neum/glopt/moretest More test set]<br />
* [http://www.mat.univie.ac.at/~neum/software/dixon.tar.gz Dixon-Szegö test set]<br />
* [http://titan.princeton.edu/TestProblems/ Handbook of Test Problems for local and global optimization]<br />
<br />
=== Optimization ===<br />
<br />
* [http://www.tik.ee.ethz.ch/sop/pisa/ PISA optimization framework]<br />
* [http://www.jeo.org/emo/EMOOsoftware.html Impl. of multiobjective optimization methods]<br />
* [http://www.hvass-labs.org/projects/swarmops/ swarmOps]<br />
* [http://www.icsi.berkeley.edu/~storn/code.html Differential Evolution]<br />
* [http://control.ee.ethz.ch/~joloef/wiki/pmwiki.php YALMIP] High-level optimization problem solver, uses external solvers as backend<br />
* [http://www.mat.univie.ac.at/~neum/glopt/software_g.html Global optimization software]<br />
* [http://www.gerad.ca/NOMAD/Abramson/nomadm.html Mesh-Adaptive Direct Search (MADS) software in Matlab]<br />
* [http://www-rocq.inria.fr/~gilbert/modulopt/modulopt.html Modulopt]: fortran implementations and some matlab<br />
* [http://www.geodise.org/documentation/OptionsMatlab/html/ MatlabOptions] Matlab interface to the design and optimization package (OPTIONS).<br />
<br />
=== Various links ===<br />
<br />
* [http://videolectures.net/Top/Computer_Science/Machine_Learning/ An excellent collection of machine learning lecture videos]<br />
* [http://home.online.no/~pjacklam/matlab/software/util/fullindex.html A collection of useful Matlab scripts]<br />
* [http://home.online.no/~pjacklam/matlab/doc/mtt/ MATLAB array manipulation tips and tricks]<br />
* [http://www.adaptivebox.net/research/bookmark/psocodes_link.html Particle Swarm implementations]<br />
* [http://mloss.org A great selection of open source machine learning software]<br />
* [http://www.cimlcommunity.org/ Another repository of machine learning tools]<br />
* [http://stommel.tamu.edu/~baum/toolboxes.html A list of Matlab toolboxes]<br />
* [http://mdoboard.proboards59.com/ ISSMO-REASON: Research and Engineering Applications in Structural Optimization Network]<br />
* [http://www.nafems.org/about/ NAFEMS]<br />
* [http://www.kat-net.net/ The European Coordinating Action on Key Aerodynamic Technologies]<br />
* [http://www.altairhyperworks.co.uk/Default.aspx Altair Hyperworks]<br />
* [http://www.ifte.de/english/research/index.html IFTE]<br />
* [http://www.optiy.eu/Features.html OptyI]<br />
* [http://www.technet-alliance.com/ Technet Alliance]<br />
<br />
=== Conference links ===<br />
<br />
* [http://www.conferencealerts.com Conference alerts]<br />
* [http://www.conferencealerts.com/engineer.htm Engineering conferences]<br />
* [http://www.conferencealerts.com/ai.htm AI Conference alerts]<br />
* [http://www.ieee.org/web/conferences/search/index.html IEEE conferences]<br />
* [http://www.aiaa.org/content.cfm?pageid=1 AIAA Conferences]<br />
* [http://www.wikicfp.com/ Wiki of conferences]<br />
* [http://ddl.me.cmu.edu/ddwiki/index.php/ASME_IDETC/CIE_Conferences CIE Conferences]<br />
* [http://users.jyu.fi/~miettine/lista.html#Conferences Kaisa's conferences]<br />
* [http://www.conference-service.com/conferences/neural-networks.html AI Conferences]<br />
* [http://ieee-cis.org/conferences/co_sponsorship_1/ IEEE CIS conferences]<br />
* [http://www.makhfi.com/events.htm Neural Network Events]<br />
* [http://www.adam.ntu.edu.sg/~mgeorg/lately_announced.pl?timeback=30 Neural Network Conferences]<br />
* [http://openresearch.org/mw/index.php?title=Upcoming_deadlines&field=Machine+learning Open research AI conferences]<br />
<br />
<br />
=== Journal Links ===<br />
<br />
*[http://www.linklings.net/tomacs/charter.html ACM Transactions on Modeling and Computer Simulation]<br />
*[http://www.elsevier.com/wps/find/journaldescription.cws_home/622330/description#description Simulation Modelling Practice and Theory]<br />
*[http://www.elsevier.com/wps/find/journaldescription.cws_home/422911/description#description Advances in Engineering Software]<br />
*[http://journaltool.asme.org/Content/JournalDescriptions.cfm?journalId=12 Journal of Mechanical Design ]<br />
*[http://journaltool.asme.org/Content/JournalDescriptions.cfm?journalId=3&Journal=JCISE Journal of Computing and Information Science in Engineering]<br />
*[http://www.elsevier.com/wps/find/journaldescription.cws_home/975/description#description Engineering Applications of Artificial Intelligence]<br />
*[http://www.elsevier.com/wps/find/journaldescription.cws_home/622240/description#description Advanced Engineering Informatics]<br />
*[http://www.springer.com/computer/information+systems/journal/366 Engineering with Computers]<br />
*[http://www.springer.com/computer/mathematics/journal/521 Neural Computing and Applications]<br />
*[http://www.aiaa.org/content.cfm?pageid=322&lupubid=2 AIAA Journal]<br />
*[http://journals.cambridge.org/action/displayJournal?jid=aie Artificial Intelligence for Engineering Design, Analysis and Manufacturing ]<br />
*[http://jmlr.csail.mit.edu/ Journal of Machine Learning Research]<br />
*[http://www.elsevier.com/wps/find/journaldescription.cws_home/505645/description#description Computer Methods in Applied Mechanics and Engineering]<br />
*[http://www.elsevier.com/wps/find/journaldescription.cws_home/524998/description#description Applied Mathematical Modelling]<br />
*[http://www.siam.org/journals/sisc.php SIAM Journal on Scientific Computing]<br />
*[http://www.iop.org/EJ/journal/-page=scope/0266-5611 Inverse Problems]<br />
*[http://www.tandf.co.uk/journals/titles/17415977.asp Inverse problems in science and engineering]<br />
*[http://ieeexplore.ieee.org/xpl/RecentIssue.jsp?punumber=3468 IEEE Transactions on Systems, Man and Cybernetics, Part A]<br />
*[http://ieeexplore.ieee.org/xpl/RecentIssue.jsp?punumber=3477 IEEE Transactions on Systems, Man and Cybernetics, Part B]<br />
*[http://ieeexplore.ieee.org/xpl/RecentIssue.jsp?punumber=5326 IEEE Transactions on Systems, Man and Cybernetics, Part C]<br />
*[http://www.techscience.com/cmes/aims_scope.html Computer Modeling in Engineering & Sciences]<br />
*[http://www.informaworld.com/smpp/title~db=all~content=t713723652~tab=summary Journal of Experimental & Theoretical Artificial Intelligence]<br />
*[http://www.elsevier.com/locate/jocs Journal of Computational Science]</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=Config:InitialDesign&diff=5112Config:InitialDesign2010-03-25T10:40:27Z<p>Icouckuy: AutoConfig for SUMO 6.2</p>
<hr />
<div>'''Generated for SUMO toolbox version 7.0'''.<br />
''We are well aware that documentation is not always complete and possibly even out of date in some cases. We try to document everything as best we can but much is limited by available time and manpower. We are are a university research group after all. The most up to date documentation can always be found (if not here) in the default.xml configuration file and, of course, in the source files. If something is unclear please dont hesitate to [[Reporting problems|ask]].''<br />
== InitialDesign ==<br />
<br />
=== lhd ===<br />
Latin Hypercube DOE<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:InitialDesign|InitialDesign]] type="[[InitialDesign#LatinHypercubeDesign|LatinHypercubeDesign]]"><br />
<!-- how many points to generate --><br />
<Option key="points" value="20"/><br />
<!--<Option key="weight" value="0.5"/>--><br />
<!--<Option key="coolingFactor" value="0.9"/>--><br />
<!--<Option key="p" value="5.0"/>--><br />
</[[Config:InitialDesign|InitialDesign]]><br />
</source><br />
=== factorial ===<br />
Specifies a simple Factorial Design (uniform grid)<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:InitialDesign|InitialDesign]] type="[[InitialDesign#FactorialDesign|FactorialDesign]]"><br />
<!-- how many points to generate for each dimension as a vector --><br />
<!-- a scalar value (l) is the same as [l l ... l] (length of input dimension) --><br />
<Option key="levels" value="3"/><br />
</[[Config:InitialDesign|InitialDesign]]><br />
</source><br />
=== random ===<br />
Specifies a trivial Random design<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:InitialDesign|InitialDesign]] type="[[InitialDesign#RandomDesign|RandomDesign]]"><br />
<Option key="points" value="20"/><br />
</[[Config:InitialDesign|InitialDesign]]><br />
</source><br />
=== lhdWithCornerPoints ===<br />
Specifies a combined Latin HyperCube and FactorialDesign<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:InitialDesign|InitialDesign]] type="[[InitialDesign#CombinedDesign|CombinedDesign]]"><br />
<!-- Select samples in a Latin Hypercube Design --><br />
<[[Config:InitialDesign|InitialDesign]] type="[[InitialDesign#LatinHypercubeDesign|LatinHypercubeDesign]]"><br />
<!-- how many points to generate --><br />
<Option key="points" value="20"/><br />
<!--<Option key="weight" value="0.5"/>--><br />
<!--<Option key="coolingFactor" value="0.9"/>--><br />
<!--<Option key="p" value="5.0"/>--><br />
</[[Config:InitialDesign|InitialDesign]]><br />
<br />
<[[Config:InitialDesign|InitialDesign]] type="[[InitialDesign#FactorialDesign|FactorialDesign]]"><br />
<!-- how many points to generate for each dimension as a vector --><br />
<!-- a scalar value (l) is the same as [l l ... l] (length of input dimension) --><br />
<Option key="levels" value="2"/><br />
</[[Config:InitialDesign|InitialDesign]]><br />
</[[Config:InitialDesign|InitialDesign]]><br />
</source><br />
=== dataset ===<br />
Use an exsiting dataset as the initial design<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:InitialDesign|InitialDesign]] type="[[InitialDesign#DatasetDesign|DatasetDesign]]"><br />
<!-- Where should we load the dataset from? --><br />
<Option key="filename" value="/path/to/your/dataset.txt"/><br />
<!-- Does the dataset also contain responses? (instead of only the inputs) <br />
If so, hasOutputs must be set to yes and the range of the inputs in the file must<br />
match the range of the simulator. If not, the range of the inputs must be [-1 1].<br />
--><br />
<Option key="hasOutputs" value="no"/><br />
</[[Config:InitialDesign|InitialDesign]]><br />
</source></div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=Config:BasisFunction&diff=5111Config:BasisFunction2010-03-25T10:40:11Z<p>Icouckuy: AutoConfig for SUMO 6.2</p>
<hr />
<div>'''Generated for SUMO toolbox version 7.0'''.<br />
''We are well aware that documentation is not always complete and possibly even out of date in some cases. We try to document everything as best we can but much is limited by available time and manpower. We are are a university research group after all. The most up to date documentation can always be found (if not here) in the default.xml configuration file and, of course, in the source files. If something is unclear please dont hesitate to [[Reporting problems|ask]].''<br />
== BasisFunction ==<br />
<br />
=== corrgauss ===<br />
Build kriging models using the maximum likelihood to set the thetas<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:BasisFunction|BasisFunction]] type="[[BasisFunction#BasisFunction|BasisFunction]]" name="corrgauss"><br />
<[[Config:Parameter|Parameter]] name="theta" min="-2" max="2" scale="log" duplicate="true"/><br />
</[[Config:BasisFunction|BasisFunction]]><br />
</source><br />
=== correxp ===<br />
Build kriging models using the maximum likelihood to set the thetas<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:BasisFunction|BasisFunction]] type="[[BasisFunction#BasisFunction|BasisFunction]]" name="correxp"><br />
<[[Config:Parameter|Parameter]] name="theta" min="-2" max="2" scale="log" duplicate="true"/><br />
</[[Config:BasisFunction|BasisFunction]]><br />
</source></div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=Config:Logging&diff=5110Config:Logging2010-03-25T10:40:02Z<p>Icouckuy: AutoConfig for SUMO 6.2</p>
<hr />
<div>'''Generated for SUMO toolbox version 7.0'''.<br />
''We are well aware that documentation is not always complete and possibly even out of date in some cases. We try to document everything as best we can but much is limited by available time and manpower. We are an university research group after all. The most up to date documentation can always be found (if not here) in the default.xml configuration file and, of course, in the source files. If something is unclear please dont hesitate to [[Reporting problems|ask]].''<br />
== Logging ==<br />
=== RootLogger ===<br />
Root logger<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<br />
<!--Root logger--><br />
<[[Config:RootLogger|RootLogger]]><br />
<br />
<!-- log all run-specific information in the main log as well --><br />
<Option key="runsInMainLog" value="true"/><br />
<br />
<!-- Set the default logging level for the root logger --><br />
<Option key="Level" value="INFO"/><br />
<br />
<!-- Specify the handlers to create in the root logger<br />
(all loggers are children of the root logger). The handlers determine<br />
where logging output is sent to. <br />
<br />
Possible levels are: OFF, SEVERE, WARNING, INFO, FINE, FINER, FINEST, ALL<br />
--> <br />
<[[Config:Handlers|Handlers]]><br />
<!-- Configure ConsoleHandler (= output to the screen) --><br />
<[[Config:ConsoleHandler|ConsoleHandler]]><br />
<Option key="Level" value="INFO"/><br />
</[[Config:ConsoleHandler|ConsoleHandler]]><br />
<br />
<!-- Configure FileHandler (= output to file)--><br />
<[[Config:FileHandler|FileHandler]]><br />
<Option key="Level" value="ALL"/><br />
</[[Config:FileHandler|FileHandler]]><br />
</[[Config:Handlers|Handlers]]><br />
</[[Config:RootLogger|RootLogger]]></source><br />
=== Custom Options ===<br />
Available options:<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<br />
<!--<Option key="loggername" value="level" />--><br />
<Option key="ibbt.sumo" value="FINEST"/><br />
<!--<Option key="loggername" value="level" />--><br />
<Option key="Matlab" value="FINEST"/></source></div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=Config:ToolboxConfiguration&diff=5109Config:ToolboxConfiguration2010-03-25T10:39:35Z<p>Icouckuy: AutoConfig for SUMO 6.2</p>
<hr />
<div>== Toolbox configuration file ==<br />
This is the default SUMO toolbox configuration, this is what gets used when you run 'go' without any arguments You can edit this file directly or make a copy and run that. See the wiki for detailed information.<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:ToolboxConfiguration|ToolboxConfiguration]] version="7.0"><br />
<[[Config:Plan|Plan]]/><br />
<[[Config:ContextConfig|ContextConfig]]/><br />
<[[Config:Logging|Logging]]/><br />
<[[Config:LevelPlot|LevelPlot]]/><br />
<[[Config:SUMO|SUMO]]/><br />
<[[Config:SampleEvaluator|SampleEvaluator]]/><br />
<[[Config:SampleSelector|SampleSelector]]/><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]/><br />
<[[Config:BasisFunction|BasisFunction]]/><br />
<[[Config:InitialDesign|InitialDesign]]/><br />
<[[Config:Optimizer|Optimizer]]/><br />
</[[Config:ToolboxConfiguration|ToolboxConfiguration]]><br />
</source></div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=Config:SequentialDesign&diff=5108Config:SequentialDesign2010-03-25T10:38:45Z<p>Icouckuy: AutoConfig for SUMO 6.2</p>
<hr />
<div>'''Generated for SUMO toolbox version 7.0'''.<br />
''We are well aware that documentation is not always complete and possibly even out of date in some cases. We try to document everything as best we can but much is limited by available time and manpower. We are are a university research group after all. The most up to date documentation can always be found (if not here) in the default.xml configuration file and, of course, in the source files. If something is unclear please dont hesitate to [[Reporting problems|ask]].''<br />
== SampleSelector ==<br />
<br />
=== empty ===<br />
Don't select any new samples, useful when modeling multiple outputs, and you don't want to involve one of these outputs in the sampling process.<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#EmptySampleSelector|EmptySampleSelector]]" combineOutputs="false"/><br />
</source><br />
=== random ===<br />
Selects new samples randomly in the design space.<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#RandomSampleSelector|RandomSampleSelector]]" combineOutputs="false"/><br />
</source><br />
=== delaunay ===<br />
This sample selector uses a Delaunay triangulation of the data to select samples in locations far from previous samples, or in locations where the estimated model error is largest. This algorithm uses QHull, which is very slow for high dimensions, so you should only use this sample selector for less than 6D and for less than 1000 samples.<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#PipelineSampleSelector|PipelineSampleSelector]]" combineOutputs="false"><br />
<br />
<[[Config:CandidateGenerator|CandidateGenerator]] type="[[CandidateGenerator#DelaunayCandidateGenerator|DelaunayCandidateGenerator]]"/><br />
<br />
<[[Config:CandidateRanker|CandidateRanker]] type="[[CandidateRanker#modelDifference|modelDifference]]"><br />
<Option key="criterion_parameter" value="2"/><br />
</[[Config:CandidateRanker|CandidateRanker]]><br />
<[[Config:CandidateRanker|CandidateRanker]] type="[[CandidateRanker#delaunayVolume|delaunayVolume]]"/><br />
<br />
<[[Config:MergeCriterion|MergeCriterion]] type="[[MergeCriterion#WeightedAverage|WeightedAverage]]" weights="[1 1]"/><br />
<br />
</[[Config:SampleSelector|SampleSelector]]><br />
</source><br />
=== density ===<br />
A space-filling sampling algorithm which uses an approximation of the Voronoi tessellation of the design space. Will only sample within the "allowed" areas if constraints are specified.<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#VoronoiSampleRanker|VoronoiSampleRanker]]" combineOutputs="false"/><br />
</source><br />
=== error ===<br />
An adaptive sample selection algorithm (error based), driven by the evaluation of your model on a dense grid, which selects samples in locations where the model error is estimated to be the largest.<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#PipelineSampleSelector|PipelineSampleSelector]]" combineOutputs="false"><br />
<br />
<[[Config:CandidateGenerator|CandidateGenerator]] type="[[CandidateGenerator#GridCandidateGenerator|GridCandidateGenerator]]"/><br />
<br />
<[[Config:CandidateRanker|CandidateRanker]] type="[[CandidateRanker#modelDifference|modelDifference]]"><br />
<Option key="criterion_parameter" value="4"/><br />
</[[Config:CandidateRanker|CandidateRanker]]><br />
<br />
<[[Config:MergeCriterion|MergeCriterion]] type="[[MergeCriterion#ClosenessThreshold|ClosenessThreshold]]"><br />
<br />
<!-- Closeness threshold, Double --><br />
<Option key="closenessThreshold" value="0.05"/><br />
<!-- Set a % of the maximumSamples to randomly chosen --><br />
<Option key="randomPercentage" value="20"/><br />
<br />
<Option key="debug" value="off"/><br />
</[[Config:MergeCriterion|MergeCriterion]]><br />
</[[Config:SampleSelector|SampleSelector]]><br />
</source><br />
=== lola-voronoi ===<br />
A highly adaptive sampling algorithm which performs a trade-off between exploration (filling up the design space as equally as possible) and exploitation (selecting data points in highly nonlinear regions). lola-voronoi is the only sample selector which currently supports multiple outputs, auto-sampled inputs and constraints.<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#LOLAVoronoiSampleSelector|LOLAVoronoiSampleSelector]]" combineOutputs="false"><br />
<!-- Integer between 2 and 20 --><br />
<Option key="neighbourhoodSize" value="2"/><br />
<!-- Number of frequency values returned for each submitted sample. Only used with auto-sampled inputs. --><br />
<Option key="frequencies" value="11"/><br />
</[[Config:SampleSelector|SampleSelector]]><br />
</source><br />
=== rationalPoleSupression ===<br />
A sampling algorithm aimed at supressing poles in rational models by sampling them (only for Rational models)<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#OptimizeCriterion|OptimizeCriterion]]" combineOutputs="false"><br />
<br />
<!-- This criterion has to be solved to choose new samples, one can choose the optimizer used here --><br />
<[[Config:Optimizer|Optimizer]]>patternsearch</[[Config:Optimizer|Optimizer]]><br />
<br />
<[[Config:CandidateRanker|CandidateRanker]] type="[[CandidateRanker#rationalPoleSupression|rationalPoleSupression]]" scaling="none"/><br />
<[[Config:CandidateRanker|CandidateRanker]] type="[[CandidateRanker#modelDifference|modelDifference]]" scaling="none"/><br />
<br />
<!--<br />
when debug is 'on' a contour plot of the criterion function is drawn every iteration.<br />
Together with the current samples and the chosen samples<br />
--><br />
<Option key="debug" value="off"/><br />
</[[Config:SampleSelector|SampleSelector]]><br />
</source><br />
=== expectedImprovement ===<br />
A sampling algorithm aimed at optimization problems (only for Kriging and RBF)<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#OptimizeCriterion|OptimizeCriterion]]" combineOutputs="false"><br />
<br />
<!-- This criterion has to be solved to choose new samples, one can choose the optimizer used here --><br />
<[[Config:Optimizer|Optimizer]]>patternsearch</[[Config:Optimizer|Optimizer]]><br />
<br />
<[[Config:CandidateRanker|CandidateRanker]] type="[[CandidateRanker#expectedImprovement|expectedImprovement]]" scaling="none"><br />
</[[Config:CandidateRanker|CandidateRanker]]><br />
<[[Config:CandidateRanker|CandidateRanker]] type="[[CandidateRanker#maxvar|maxvar]]" scaling="none"/><br />
<br />
<!--<br />
when debug is 'on' a contour plot of the criterion function is drawn every iteration.<br />
Together with the current samples and the chosen samples<br />
--><br />
<Option key="debug" value="off"/><br />
</[[Config:SampleSelector|SampleSelector]]><br />
</source><br />
=== extremaLOLA ===<br />
LOLA-Voronoi sample selector supplemented with 1 sample at the minimum and maximum<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#CombinedSampleSelector|CombinedSampleSelector]]" combineOutputs="false"><br />
<!-- A highly adaptive sampling algorithm, error and density based --><br />
<[[Config:SampleSelector|SampleSelector]] weight="0.8">lola-voronoi</[[Config:SampleSelector|SampleSelector]]><br />
<[[Config:SampleSelector|SampleSelector]] weight="0.1">sampleMinimum</[[Config:SampleSelector|SampleSelector]]><br />
<[[Config:SampleSelector|SampleSelector]] weight="0.1">sampleMaximum</[[Config:SampleSelector|SampleSelector]]><br />
<br />
<[[Config:MergeCriterion|MergeCriterion]] type="[[MergeCriterion#ClosenessThreshold|ClosenessThreshold]]"><br />
<br />
<!-- Closeness threshold, Double --><br />
<Option key="closenessThreshold" value="0.05"/><br />
<!-- Set a % of the maximumSamples to randomly chosen --><br />
<Option key="randomPercentage" value="0"/><br />
<br />
<Option key="debug" value="off"/><br />
</[[Config:MergeCriterion|MergeCriterion]]><br />
</[[Config:SampleSelector|SampleSelector]]><br />
</source><br />
=== sampleMinimum ===<br />
Selects one sample at the minimum of the model.<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#OptimizeCriterion|OptimizeCriterion]]" combineOutputs="false"><br />
<[[Config:Optimizer|Optimizer]]>patternsearch</[[Config:Optimizer|Optimizer]]><br />
<[[Config:CandidateRanker|CandidateRanker]] type="[[CandidateRanker#minmodel|minmodel]]" scaling="none"/><br />
</[[Config:SampleSelector|SampleSelector]]><br />
</source><br />
=== sampleMaximum ===<br />
Selects one sample at the maximum of the model.<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#OptimizeCriterion|OptimizeCriterion]]" combineOutputs="false"><br />
<[[Config:Optimizer|Optimizer]]>patternsearch</[[Config:Optimizer|Optimizer]]><br />
<[[Config:CandidateRanker|CandidateRanker]] type="[[CandidateRanker#maxmodel|maxmodel]]" scaling="none"/><br />
</[[Config:SampleSelector|SampleSelector]]><br />
</source><br />
=== default ===<br />
LOLA sample selector combined with error based sample selector<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:SampleSelector|SampleSelector]] type="[[SampleSelector#CombinedSampleSelector|CombinedSampleSelector]]" combineOutputs="false"><br />
<[[Config:SampleSelector|SampleSelector]] weight="0.7">lola-voronoi</[[Config:SampleSelector|SampleSelector]]><br />
<[[Config:SampleSelector|SampleSelector]] weight="0.3">error</[[Config:SampleSelector|SampleSelector]]><br />
<br />
<[[Config:MergeCriterion|MergeCriterion]] type="[[MergeCriterion#ClosenessThreshold|ClosenessThreshold]]"> <br />
<!-- Closeness threshold, Double --><br />
<Option key="closenessThreshold" value="0.05"/><br />
<!-- Set a % of the maximumSamples to randomly chosen --><br />
<Option key="randomPercentage" value="0"/><br />
<br />
<Option key="debug" value="off"/><br />
</[[Config:MergeCriterion|MergeCriterion]]><br />
</[[Config:SampleSelector|SampleSelector]]><br />
</source></div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=Config:DataSource&diff=5107Config:DataSource2010-03-25T10:38:19Z<p>Icouckuy: AutoConfig for SUMO 6.2</p>
<hr />
<div>'''Generated for SUMO toolbox version 7.0'''.<br />
''We are well aware that documentation is not always complete and possibly even out of date in some cases. We try to document everything as best we can but much is limited by available time and manpower. We are are a university research group after all. The most up to date documentation can always be found (if not here) in the default.xml configuration file and, of course, in the source files. If something is unclear please dont hesitate to [[Reporting problems|ask]].''<br />
== SampleEvaluator ==<br />
<br />
=== local ===<br />
Use this if you data generator is a native executable, shell script, or java class<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:SampleEvaluator|SampleEvaluator]] id="local" type="ibbt.sumo.sampleevaluators.LocalSampleEvaluator"><br />
<!-- Maximum number of times to resubmit a point (e.g., in case something went wrong) --><br />
<Option key="maxResubmissions" value="1"/><br />
<!-- If a sample takes longer than "sampleTimeout*average evaluation time of one sample" <br />
seconds to evaluate it is removed from the pending list (set to -1 to disable) --><br />
<Option key="sampleTimeout" value="-1"/><br />
<!-- Can be set to "java" for java executables, to "external" for platform-specific<br />
binaries/scripts and to nothing at all for auto-detection.--><br />
<Option key="simulatorType" value=""/><br />
<!-- Can be set to a number of seconds, if one simulator evaluation exceeds this timeout,<br />
the simulation is aborted --><br />
<!-- <Option key="timeout" value="12"/> --><br />
<!-- Number of samples to evaluate concurrently, useful if you have a dual or multi-core machine --><br />
<Option key="threadCount" value="1"/><br />
</[[Config:SampleEvaluator|SampleEvaluator]]><br />
</source><br />
=== matlab ===<br />
Evaluate samples using a matlab script (ie. your simulator is a matlab script)<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:SampleEvaluator|SampleEvaluator]] id="matlab" type="ibbt.sumo.sampleevaluators.matlab.MatlabSampleEvaluator"><br />
<!-- Maximum number of times to resubmit a point (e.g., in case something went wrong) --><br />
<Option key="maxResubmissions" value="1"/><br />
<!-- If a sample takes longer than "sampleTimeout*average evaluation time of one sample" <br />
seconds to evaluate it is removed from the pending list (set to -1 to disable) --><br />
<Option key="sampleTimeout" value="-1"/><br />
</[[Config:SampleEvaluator|SampleEvaluator]]><br />
</source><br />
=== griddedDataset ===<br />
Evaluate samples using a gridded dataset<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:SampleEvaluator|SampleEvaluator]] id="griddedDataset" type="ibbt.sumo.sampleevaluators.datasets.GriddedDatasetSampleEvaluator"><br />
<!-- Using an ID you can specify which dataset from the simulator file to use --><br />
<!-- <Option key="id" value="someDataset"/> --><br />
</[[Config:SampleEvaluator|SampleEvaluator]]><br />
</source><br />
=== scatteredDataset ===<br />
Evaluate samples using a scattered dataset<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:SampleEvaluator|SampleEvaluator]] id="scatteredDataset" type="ibbt.sumo.sampleevaluators.datasets.ScatteredDatasetSampleEvaluator"><br />
<!-- Using an ID you can specify which dataset from the simulator file to use --><br />
<!-- <Option key="id" value="someDataset"/> --><br />
</[[Config:SampleEvaluator|SampleEvaluator]]><br />
</source><br />
=== calcua ===<br />
Evaluate samples on a SGE administered cluster through a remote, ssh reachable frontnode<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:SampleEvaluator|SampleEvaluator]] id="calcua" type="ibbt.sumo.sampleevaluators.distributed.sge.RemoteSGESampleEvaluator"><br />
<!-- Maximum number of times to resubmit a point (e.g., in case something went wrong) --><br />
<Option key="maxResubmissions" value="1"/><br />
<!-- If a sample takes longer than "sampleTimeout*average evaluation time of one sample" <br />
seconds to evaluate it is removed from the pending list (set to -1 to disable) --><br />
<Option key="sampleTimeout" value="-1"/><br />
<br />
<!--The platform specs of the cluster--><br />
<[[Config:Executable|Executable]] platform="linux" arch="x86_64"/><br />
<[[Config:Backend|Backend]] id="remoteSGE" type="ibbt.sumo.sampleevaluators.distributed.sge.RemoteSGEBackend"><br />
<!--ssh user name to login on the front node--><br />
<Option key="user" value="dgorisse"/><br />
<!--Submissions happen from this front node, you need to have key-based ssh authentication--><br />
<Option key="frontNode" value="submit.calcua.ua.ac.be"/><br />
<!--Directory on the front node where input/output files, dependencies, etc. are stored--><br />
<Option key="remoteDirectory" value="/storeA/users/dgorisse/output"/><br />
<!--poll for result files every xx seconds--><br />
<Option key="pollInterval" value="20"/><br />
<!--queues we can submit to--><br />
<Option key="queues" value="all.q,fast.q"/><br />
<!--check for faster queues (more slots available) every xx seconds--><br />
<Option key="queueRevisionRate" value="10"/><br />
<!--script on the frontnode that sets up the necessary shell environment--><br />
<Option key="environmentCommand" value=". ~/.profile;"/><br />
</[[Config:Backend|Backend]]><br />
</[[Config:SampleEvaluator|SampleEvaluator]]><br />
</source></div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=Config:ModelBuilder&diff=5106Config:ModelBuilder2010-03-25T10:37:33Z<p>Icouckuy: AutoConfig for SUMO 6.2</p>
<hr />
<div>'''Generated for SUMO toolbox version 7.0'''.<br />
''We are well aware that documentation is not always complete and possibly even out of date in some cases. We try to document everything as best we can but much is limited by available time and manpower. We are are a university research group after all. The most up to date documentation can always be found (if not here) in the default.xml configuration file and, of course, in the source files. If something is unclear please dont hesitate to [[Reporting problems|ask]].''<br />
== AdaptiveModelBuilder ==<br />
<br />
=== rational ===<br />
Build rational models using a custom stochastic hillclimber to select the model parameters<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#SequentialModelBuilder|SequentialModelBuilder]]" combineOutputs="true"><br />
<!-- Maximum number of models built before selecting new samples --> <br />
<Option key="maximumRunLength" value="30"/><br />
<!-- Degeneration of score if a model gets older --><br />
<Option key="decay" value=".99"/><br />
<!-- Size of the best model history --><br />
<Option key="historySize" value="15"/><br />
<!-- One of best, last. When set to best the best `historySize' models are kept,<br />
- - when set to last, the last `historySize' models are kept --><br />
<Option key="strategy" value="best"/><br />
<!-- <Option key="strategy" value="window"/> --><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#RationalFactory|RationalFactory]]"><br />
<!-- Bounds for the weights of the rational modeller --><br />
<Option key="weightBounds" value="1,40"/><br />
<!-- Bounds for the percentage of degrees of freedom wrt number of samples --><br />
<Option key="percentBounds" value="1,100"/><br />
<!-- Regardless of the percentage bounds, never use more than this many degrees of freedom --><br />
<Option key="maxDegrees" value="80"/><br />
<!-- When randomizing rational flags, what percentage should be set --><br />
<Option key="percentRational" value="70"/><br />
<!-- If a variable is named "f" of "frequency" it will be modelled differently, if this is set to auto,<br />
- - If this field is set to a variable name, that variable will be considered to be the frequency --><br />
<Option key="frequencyVariable" value="auto"/><br />
<!-- Base function for interpolation, one of chebyshev, power, legendre --><br />
<Option key="basis" value="power"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== rationalgenetic ===<br />
Build rational models using a genetic algorithm<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#GeneticModelBuilder|GeneticModelBuilder]]" combineOutputs="false"><br />
<!--See that matlab gads toolbox documentation for more information on the options--><br />
<Option key="restartStrategy" value="continue"/><br />
<Option key="populationType" value="custom"/><br />
<Option key="populationSize" value="30"/><br />
<Option key="crossoverFraction" value="0.7"/><br />
<Option key="maxGenerations" value="20"/><br />
<Option key="eliteCount" value="1"/><br />
<Option key="stallGenLimit" value="5"/><br />
<Option key="stallTimeLimit" value="Inf"/><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#RationalFactory|RationalFactory]]"><br />
<Option key="crossoverFcn" value="crossover"/><br />
<Option key="mutationFcn" value="mutation"/><br />
<Option key="creationFcn" value="createInitialPopulation"/><br />
<!-- Use the next three functions instead of the previous three if you set the<br />
population type to doubleVector --><br />
<!--<Option key="creationFcn" value="@gacreationuniform"/><br />
<Option key="crossoverFcn" value="@crossoverheuristic"/><br />
<Option key="mutationFcn" value="@mutationadaptfeasible"/>--><br />
<br />
<!-- Bounds for the weights of the rational modeller --><br />
<Option key="weightBounds" value="1,40"/><br />
<!-- Bounds for the percentage of degrees of freedom wrt number of samples --><br />
<Option key="percentBounds" value="1,100"/><br />
<!-- Regardless of the percentage bounds, never use more than this many degrees of freedom --><br />
<Option key="maxDegrees" value="80"/><br />
<!-- When randomizing rational flags, what percentage should be set --><br />
<Option key="percentRational" value="70"/><br />
<!-- If a variable is named "f" of "frequency" <br />
it will be modelled differently, if this is set to auto --><br />
<!-- If this field is set to a variable name, that variable will be considered to be the frequency --><br />
<Option key="frequencyVariable" value="auto"/><br />
<!-- Base function for interpolation, one of chebyshev, power, legendre --><br />
<Option key="basis" value="power"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== rationalpso ===<br />
Generate Rational models using PSO<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#OptimizerModelBuilder|OptimizerModelBuilder]]" combineOutputs="false"><br />
<!-- Plot the optimization surface, visualizes the search through the parameter space (2D only) --><br />
<Option key="plotOptimSurface" value="false"/><br />
<!-- Re-start strategy for resuming the optimization process between sampling iterations.<br />
One of 'random','continue','model' and 'intelligent' (Default). See the docs for more information --><br />
<Option key="restartStrategy" value="intelligent"/><br />
<br />
<[[Config:Optimizer|Optimizer]] type="[[Optimizer#PSOtOptimizer|PSOtOptimizer]]"><br />
<Option key="typePSO" value="0"/><br />
<Option key="seedPSO" value="1"/><br />
<Option key="popSize" value="10"/><br />
<Option key="maxiters" value="10"/><br />
<Option key="epochInertia" value="8"/><br />
<Option key="gradientTermination" value="8"/><br />
</[[Config:Optimizer|Optimizer]]><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#RationalFactory|RationalFactory]]"><br />
<!-- Bounds for the weights of the rational modeller --><br />
<Option key="weightBounds" value="1,40"/><br />
<!-- Bounds for the percentage of degrees of freedom wrt number of samples --><br />
<Option key="percentBounds" value="1,100"/><br />
<!-- Regardless of the percentage bounds, never use more than this many degrees of freedom --><br />
<Option key="maxDegrees" value="80"/><br />
<!-- When randomizing rational flags, what percentage should be set --><br />
<Option key="percentRational" value="70"/><br />
<!-- If a variable is named "f" of "frequency" <br />
it will be modelled differently, if this is set to auto --><br />
<!-- If this field is set to a variable name, that variable will be considered to be the frequency --><br />
<Option key="frequencyVariable" value="auto"/><br />
<!-- Base function for interpolation, one of chebyshev, power, legendre --><br />
<Option key="basis" value="power"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== polynomialfixed ===<br />
Build polynomial models with a fixed order<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#AdaptiveModelBuilder|AdaptiveModelBuilder]]" combineOutputs="false"><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#PolynomialFactory|PolynomialFactory]]"><br />
<!-- Specifies the structure of the polynomial --><br />
<!-- Expects matlab matrix: Element (i,j) is the exponent of variable i for term j. --><br />
<Option key="degrees" value="[0 0;1 0;0 1;1 1;2 0;0 2;2 2]"/> <!-- Equals to 1+x+y+xy+xx+yy+xxyy --><br />
<!-- Base function for interpolation, one of chebyshev, power, legendre --><br />
<Option key="basis" value="power"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== rbf ===<br />
Build Radial Basis Function models<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#SequentialModelBuilder|SequentialModelBuilder]]" combineOutputs="false"><br />
<!-- Maximum number of models built before selecting new samples --> <br />
<Option key="maximumRunLength" value="20"/><br />
<!-- Degeneration of score if a model gets older --><br />
<Option key="decay" value=".9"/><br />
<!-- Size of the best model history --><br />
<Option key="historySize" value="15"/><br />
<!-- One of best, last. When set to best the best `historySize' models are kept,<br />
- - when set to last, the last `historySize' models are kept --><br />
<Option key="strategy" value="best"/><br />
<br />
<!-- <Option key="strategy" value="window"/> --><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#BFFactory|BFFactory]]"><br />
<Option key="type" value="RBF"/><br />
<br />
<[[Config:BasisFunction|BasisFunction]] name="gaussian" min=".1" max="5" scale="ln"/><br />
<[[Config:BasisFunction|BasisFunction]] name="multiquadric" min=".1" max="5" scale="ln"/><br />
<!--<[[Config:BasisFunction|BasisFunction]] name="biharmonic" min=".1" max="5" scale="ln"/> --><br />
<[[Config:BasisFunction|BasisFunction]] name="exponential" min=".1,.5" max="5,2" scale="ln,lin"/><br />
<br />
<Option key="regression" value="-1,0,1,2"/><br />
<Option key="backend" value="AP"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== rbfgenetic ===<br />
Build Radial Basis Function models using a genetic algorithm<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#GeneticModelBuilder|GeneticModelBuilder]]" combineOutputs="false"><br />
<Option key="restartStrategy" value="continue"/><br />
<!--See that matlab gads toolbox documentation for more information on the options--><br />
<Option key="populationType" value="custom"/><br />
<Option key="populationSize" value="15"/><br />
<Option key="crossoverFraction" value="0.7"/><br />
<Option key="maxGenerations" value="10"/><br />
<Option key="eliteCount" value="1"/><br />
<Option key="stallGenLimit" value="4"/><br />
<Option key="stallTimeLimit" value="Inf"/><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#BFFactory|BFFactory]]"><br />
<Option key="type" value="RBF"/><br />
<br />
<Option key="crossoverFcn" value="crossover"/><br />
<Option key="mutationFcn" value="mutation"/><br />
<Option key="creationFcn" value="createInitialPopulation"/><br />
<br />
<!-- Bounds for the shape parameter --><br />
<[[Config:BasisFunction|BasisFunction]] name="gaussian" min=".1" max="5" scale="ln"/><br />
<[[Config:BasisFunction|BasisFunction]] name="multiquadric" min=".1" max="5" scale="ln"/><br />
<!-- <[[Config:BasisFunction|BasisFunction]] name="biharmonic" min=".1" max="5" scale="ln"/> --><br />
<[[Config:BasisFunction|BasisFunction]] name="exponential" min=".1,.5" max="5,2" scale="ln,lin"/><br />
<br />
<Option key="regression" value="-1,0,1,2"/><br />
<!-- Basisfunction, one of 'multiquadric', 'triharmonic', 'biharmonic' --><br />
<!-- Specify which implementation to use, currently, 'Direct', 'AP', 'Greedy' and<br />
'FastRBF' are supported.<br />
<br />
'Direct' solves the direct problem by inverting the interpolation<br />
matrix<br />
'AP' uses an alternating projections method when the system gets<br />
too large. This is *MUCH* slower than 'Direct', and doesn't<br />
guarantee convergence, use with caution<br />
'Greedy' uses a one point greedy algorithm for selecting the <br />
interpolation centers. Same remark applies as with 'AP'<br />
'FastRBF' interfaces the FastRBF library. When using FastRBF, <br />
make sure your copy of the software is installed under <br />
the src/matlab/contrib directory and that the software <br />
is licensed properly.<br />
The FastRBF matlab toolbox can be found at<br />
http://www.farfieldtechnology.com<br />
--><br />
<Option key="backend" value="AP"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== dace ===<br />
Build DACE models (equivalent to Kriging but a custom implementation)<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#SequentialModelBuilder|SequentialModelBuilder]]" combineOutputs="false"><br />
<!-- Maximum number of models built before selecting new samples --> <br />
<Option key="maximumRunLength" value="20"/><br />
<!-- Degeneration of score if a model gets older --><br />
<Option key="decay" value=".9"/><br />
<!-- Size of the best model history --><br />
<Option key="historySize" value="15"/><br />
<!-- One of best, last. When set to best the best `historySize' models are kept,<br />
- - when set to last, the last `historySize' models are kept --><br />
<Option key="strategy" value="best"/><br />
<br />
<!-- <Option key="strategy" value="window"/> --><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#BFFactory|BFFactory]]"><br />
<Option key="type" value="DACE"/><br />
<br />
<[[Config:BasisFunction|BasisFunction]] name="gaussian" min=".1" max="5" scale="ln"/><br />
<[[Config:BasisFunction|BasisFunction]] name="multiquadric" min=".1" max="5" scale="ln"/><br />
<!--<[[Config:BasisFunction|BasisFunction]] name="biharmonic" min=".1" max="5" scale="ln"/> --><br />
<[[Config:BasisFunction|BasisFunction]] name="exponential" min=".1,.5" max="5,2" scale="ln,lin"/><br />
<br />
<Option key="regression" value="-1,0,1,2"/><br />
<Option key="backend" value="AP"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== dacegenetic ===<br />
Build DACE models using a genetic algorithm<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#GeneticModelBuilder|GeneticModelBuilder]]" combineOutputs="false"><br />
<!--See that matlab gads toolbox documentation for more information on the options--><br />
<Option key="populationType" value="custom"/><br />
<Option key="populationSize" value="15"/><br />
<Option key="crossoverFraction" value="0.7"/><br />
<Option key="maxGenerations" value="10"/><br />
<Option key="eliteCount" value="1"/><br />
<Option key="stallGenLimit" value="4"/><br />
<Option key="stallTimeLimit" value="Inf"/><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#BFFactory|BFFactory]]"><br />
<Option key="type" value="DACE"/><br />
<br />
<Option key="crossoverFcn" value="crossover"/><br />
<Option key="mutationFcn" value="mutation"/><br />
<Option key="creationFcn" value="createInitialPopulation"/><br />
<br />
<!-- Bounds for the shape parameter --><br />
<[[Config:BasisFunction|BasisFunction]] name="gaussian" min=".1" max="5" scale="ln"/><br />
<[[Config:BasisFunction|BasisFunction]] name="multiquadric" min=".1" max="5" scale="ln"/><br />
<!-- <[[Config:BasisFunction|BasisFunction]] name="biharmonic" min=".1" max="5" scale="ln"/> --><br />
<[[Config:BasisFunction|BasisFunction]] name="exponential" min=".1,.5" max="5,2" scale="ln,lin"/><br />
<br />
<Option key="regression" value="-1,0,1,2"/><br />
<Option key="backend" value="AP"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== daceps ===<br />
Build DACE using pattern search<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#OptimizerModelBuilder|OptimizerModelBuilder]]" combineOutputs="false"><br />
<!-- Plot the optimization surface, visualizes the search through the parameter space (2D only) --><br />
<Option key="plotOptimSurface" value="false"/><br />
<!-- Re-start strategy for resuming the optimization process between sampling iterations.<br />
One of 'random','continue','model' and 'intelligent' (Default). See the docs for more information --><br />
<Option key="restartStrategy" value="intelligent"/><br />
<br />
<[[Config:Optimizer|Optimizer]] type="[[Optimizer#MatlabPatternSearch|MatlabPatternSearch]]"><br />
<Option key="maxIterations" value="500"/><br />
<Option key="maxFunEvals" value="100"/><br />
<Option key="searchMethod" value="GPSPositiveBasis2N"/><br />
<Option key="pollMethod" value="MADSPositiveBasis2N"/> <br />
</[[Config:Optimizer|Optimizer]]><br />
<br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#BFFactory|BFFactory]]"><br />
<Option key="type" value="DACE"/><br />
<br />
<!--Option key="multipleBasisFunctionsAllowed" value="false"/--><br />
<br />
<[[Config:BasisFunction|BasisFunction]] name="gaussian" min=".1" max="5" scale="ln"/><br />
<br />
<Option key="regression" value="-1,0,1,2"/><br />
<Option key="backend" value="AP"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== dacepso ===<br />
Build DACE models using PSO<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#OptimizerModelBuilder|OptimizerModelBuilder]]" combineOutputs="false"><br />
<!-- Plot the optimization surface, visualizes the search through the parameter space (2D only) --><br />
<Option key="plotOptimSurface" value="false"/><br />
<!-- Re-start strategy for resuming the optimization process between sampling iterations.<br />
One of 'random','continue','model' and 'intelligent' (Default). See the docs for more information --><br />
<Option key="restartStrategy" value="intelligent"/><br />
<br />
<[[Config:Optimizer|Optimizer]] type="[[Optimizer#PSOtOptimizer|PSOtOptimizer]]"><br />
<Option key="typePSO" value="0"/><br />
<Option key="seedPSO" value="1"/><br />
<Option key="popSize" value="10"/><br />
<Option key="maxiters" value="10"/><br />
<Option key="epochInertia" value="8"/><br />
<Option key="gradientTermination" value="8"/><br />
</[[Config:Optimizer|Optimizer]]><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#BFFactory|BFFactory]]"><br />
<Option key="type" value="DACE"/><br />
<br />
<!--Option key="multipleBasisFunctionsAllowed" value="false"/--><br />
<br />
<[[Config:BasisFunction|BasisFunction]] name="gaussian" min=".1" max="5" scale="ln"/><br />
<br />
<Option key="regression" value="-1,0,1,2"/><br />
<Option key="backend" value="AP"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== gpps ===<br />
Build GP using pattern search<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#OptimizerModelBuilder|OptimizerModelBuilder]]" combineOutputs="false"><br />
<!-- Plot the optimization surface, visualizes the search through the parameter space (2D only) --><br />
<Option key="plotOptimSurface" value="false"/><br />
<!-- Re-start strategy for resuming the optimization process between sampling iterations.<br />
One of 'random','continue','model' and 'intelligent' (Default). See the docs for more information --><br />
<Option key="restartStrategy" value="intelligent"/><br />
<br />
<[[Config:Optimizer|Optimizer]] type="[[Optimizer#MatlabPatternSearch|MatlabPatternSearch]]"><br />
<Option key="maxIterations" value="500"/><br />
<Option key="maxFunEvals" value="100"/><br />
<Option key="searchMethod" value="GPSPositiveBasis2N"/><br />
<Option key="pollMethod" value="MADSPositiveBasis2N"/> <br />
</[[Config:Optimizer|Optimizer]]><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#GaussianProcessFactory|GaussianProcessFactory]]"><br />
<Option key="covFunction" value="covSEiso"/> <br />
<br />
<Option key="lowerThetaBound" value="-5"/><br />
<Option key="upperThetaBound" value="3"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== gpgenetic ===<br />
Build Gaussian Process models using a GA<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#GeneticModelBuilder|GeneticModelBuilder]]" combineOutputs="false"><br />
<!--See that matlab gads toolbox documentation for more information on the options--><br />
<Option key="populationType" value="doubleVector"/><br />
<Option key="populationSize" value="15"/><br />
<Option key="crossoverFraction" value="0.7"/><br />
<Option key="maxGenerations" value="10"/><br />
<Option key="eliteCount" value="1"/><br />
<Option key="stallGenLimit" value="4"/><br />
<Option key="stallTimeLimit" value="Inf"/><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#GaussianProcessFactory|GaussianProcessFactory]]"><br />
<Option key="creationFcn" value="@gacreationuniform"/><br />
<Option key="crossoverFcn" value="@crossoverheuristic"/><br />
<Option key="mutationFcn" value="@mutationadaptfeasible"/><br />
<!--<br />
<Option key="crossoverFcn" value="crossover"/><br />
<Option key="mutationFcn" value="mutation"/><br />
<Option key="creationFcn" value="createInitialPopulation"/><br />
--><br />
<br />
<Option key="covFunction" value="covSEiso"/><br />
<Option key="lowerThetaBound" value="-5"/><br />
<Option key="upperThetaBound" value="3"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== krigingsim ===<br />
Build kriging models using Simulated Annealing<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#OptimizerModelBuilder|OptimizerModelBuilder]]" combineOutputs="false"><br />
<!-- Plot the optimization surface, visualizes the search through the parameter space (2D only) --><br />
<Option key="plotOptimSurface" value="false"/><br />
<!-- Re-start strategy for resuming the optimization process between sampling iterations.<br />
One of 'random','continue','model' and 'intelligent' (Default). See the docs for more information --><br />
<Option key="restartStrategy" value="intelligent"/><br />
<br />
<[[Config:Optimizer|Optimizer]] type="[[Optimizer#MatlabSimAnnealing|MatlabSimAnnealing]]"><br />
<Option key="maxIterations" value="100"/><br />
<Option key="maxFunEvals" value="20"/><br />
</[[Config:Optimizer|Optimizer]]><br />
<br />
<!-- See the documentaion for possible regression and correlation functions --><br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#KrigingFactory|KrigingFactory]]"><br />
<Option key="regressionMetric" value=""/><br />
<Option key="regressionFunction" value="regpoly1"/><br />
<Option key="multipleBasisFunctionsAllowed" value="false"/><br />
<[[Config:BasisFunction|BasisFunction]]>corrgauss</[[Config:BasisFunction|BasisFunction]]><br />
</[[Config:ModelFactory|ModelFactory]]><br />
<br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== krigingps ===<br />
Build kriging models using pattern search<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#OptimizerModelBuilder|OptimizerModelBuilder]]" combineOutputs="false"><br />
<!-- Plot the optimization surface, visualizes the search through the parameter space (2D only) --><br />
<Option key="plotOptimSurface" value="false"/><br />
<!-- Re-start strategy for resuming the optimization process between sampling iterations.<br />
One of 'random','continue','model' and 'intelligent' (Default). See the docs for more information --><br />
<Option key="restartStrategy" value="intelligent"/><br />
<br />
<[[Config:Optimizer|Optimizer]] type="[[Optimizer#MatlabPatternSearch|MatlabPatternSearch]]"><br />
<Option key="maxIterations" value="100"/><br />
<Option key="maxFunEvals" value="20"/><br />
<Option key="searchMethod" value="GPSPositiveBasis2N"/><br />
<Option key="pollMethod" value="MADSPositiveBasis2N"/> <br />
</[[Config:Optimizer|Optimizer]]><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#KrigingFactory|KrigingFactory]]"><br />
<Option key="regressionMetric" value=""/><br />
<Option key="regressionFunction" value="regpoly1"/><br />
<Option key="multipleBasisFunctionsAllowed" value="true"/><br />
<[[Config:BasisFunction|BasisFunction]]>corrgauss</[[Config:BasisFunction|BasisFunction]]><br />
<[[Config:BasisFunction|BasisFunction]]>correxp</[[Config:BasisFunction|BasisFunction]]><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== krigingoptim ===<br />
Build kriging models using the matlab optimization toolbox<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#OptimizerModelBuilder|OptimizerModelBuilder]]" combineOutputs="false"><br />
<!-- Plot the optimization surface, visualizes the search through the parameter space (2D only) --><br />
<Option key="plotOptimSurface" value="false"/><br />
<!-- Re-start strategy for resuming the optimization process between sampling iterations.<br />
One of 'random','continue','model' and 'intelligent' (Default). See the docs for more information --><br />
<Option key="restartStrategy" value="intelligent"/><br />
<br />
<[[Config:Optimizer|Optimizer]] type="[[Optimizer#MatlabOptimizer|MatlabOptimizer]]"><br />
<Option key="maxIterations" value="100"/><br />
<Option key="maxFunEvals" value="20"/><br />
</[[Config:Optimizer|Optimizer]]><br />
<br />
<!-- See the documentaion for possible regression and correlation functions --><br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#KrigingFactory|KrigingFactory]]"><br />
<Option key="regressionMetric" value=""/><br />
<Option key="regressionFunction" value="regpoly1"/><br />
<Option key="multipleBasisFunctionsAllowed" value="false"/><br />
<[[Config:BasisFunction|BasisFunction]]>corrgauss</[[Config:BasisFunction|BasisFunction]]><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== kriginggenetic ===<br />
Build kriging models using a genetic algorithm<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#GeneticModelBuilder|GeneticModelBuilder]]" combineOutputs="false"><br />
<!-- Plot the optimization surface, visualizes the search through the parameter space (2D only) --><br />
<Option key="plotOptimSurface" value="false"/><br />
<!-- Re-start strategy for resuming the optimization process between sampling iterations.<br />
One of 'random','continue','model' and 'intelligent' (Default). See the docs for more information --><br />
<Option key="restartStrategy" value="intelligent"/><br />
<!-- If you specify "custom" as the population type you will be evolving models<br />
and will use the genetic operators defined in the KrigingFactory class --><br />
<Option key="populationType" value="doubleVector"/><br />
<Option key="populationSize" value="10"/><br />
<Option key="maxGenerations" value="10"/><br />
<Option key="eliteCount" value="1"/><br />
<Option key="crossoverFraction" value="0.7"/><br />
<Option key="stallGenLimit" value="4"/><br />
<Option key="stallTimeLimit" value="Inf"/><br />
<br />
<!-- See the documentaion for possible regression and correlation functions --><br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#KrigingFactory|KrigingFactory]]"><br />
<Option key="creationFcn" value="@gacreationuniform"/><br />
<Option key="crossoverFcn" value="@crossoverheuristic"/><br />
<Option key="mutationFcn" value="@mutationadaptfeasible"/><br />
<!-- <Option key="creationFcn" value="createInitialPopulation"/><br />
<Option key="crossoverFcn" value="crossover"/><br />
<Option key="mutationFcn" value="mutation"/> --><br />
<!-- See the documentaion for possible regression and correlation functions --> <Option key="regressionMetric" value=""/><br />
<Option key="regressionFunction" value="regpoly1"/><br />
<Option key="multipleBasisFunctionsAllowed" value="false"/><br />
<[[Config:BasisFunction|BasisFunction]]>corrgauss</[[Config:BasisFunction|BasisFunction]]><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== krigingpso ===<br />
Build kriging models using PSO<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#OptimizerModelBuilder|OptimizerModelBuilder]]" combineOutputs="false"><br />
<!-- Plot the optimization surface, visualizes the search through the parameter space (2D only) --><br />
<Option key="plotOptimSurface" value="false"/><br />
<!-- Re-start strategy for resuming the optimization process between sampling iterations.<br />
One of 'random','continue','model' and 'intelligent' (Default). See the docs for more information --><br />
<Option key="restartStrategy" value="intelligent"/><br />
<br />
<[[Config:Optimizer|Optimizer]] type="[[Optimizer#PSOtOptimizer|PSOtOptimizer]]"><br />
<Option key="typePSO" value="0"/><br />
<Option key="seedPSO" value="1"/><br />
<Option key="popSize" value="10"/><br />
<Option key="maxiters" value="10"/><br />
<Option key="epochInertia" value="8"/><br />
<Option key="gradientTermination" value="8"/><br />
</[[Config:Optimizer|Optimizer]]><br />
<br />
<!-- See the documentaion for possible regression and correlation functions --><br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#KrigingFactory|KrigingFactory]]"><br />
<Option key="regressionMetric" value=""/><br />
<Option key="regressionFunction" value="regpoly1"/><br />
<Option key="multipleBasisFunctionsAllowed" value="false"/><br />
<[[Config:BasisFunction|BasisFunction]]>corrgauss</[[Config:BasisFunction|BasisFunction]]><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== krigingnsga ===<br />
Build kriging models using NSGA-II, requires a multi-output or multi-measure setup<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#ParetoModelBuilder|ParetoModelBuilder]]" combineOutputs="true"><br />
<Option key="restartStrategy" value="model"/><br />
<Option key="populationSize" value="30"/><br />
<Option key="maxGenerations" value="30"/><br />
<Option key="plotParetoFront" value="false"/><br />
<Option key="paretoMode" value="true"/><br />
<br />
<!-- See the documentaion for possible regression and correlation functions --><br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#KrigingFactory|KrigingFactory]]"><br />
<Option key="regressionMetric" value=""/><br />
<Option key="regressionFunction" value="regpoly1"/><br />
<Option key="multipleBasisFunctionsAllowed" value="false"/><br />
<[[Config:BasisFunction|BasisFunction]]>corrgauss</[[Config:BasisFunction|BasisFunction]]><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== krigingrandom ===<br />
Build kriging models randomly, useful as a baseline comparison<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#RandomModelBuilder|RandomModelBuilder]]" combineOutputs="false"><br />
<!-- Plot the optimization surface, visualizes the search through the parameter space (2D only) --><br />
<Option key="plotOptimSurface" value="false"/><br />
<!-- Build 100 random models before restarting --><br />
<Option key="runSize" value="100"/><br />
<br />
<!-- See the documentaion for possible regression and correlation functions --><br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#KrigingFactory|KrigingFactory]]"><br />
<Option key="regressionMetric" value=""/><br />
<Option key="regressionFunction" value="regpoly1"/><br />
<Option key="multipleBasisFunctionsAllowed" value="false"/><br />
<[[Config:BasisFunction|BasisFunction]]>corrgauss</[[Config:BasisFunction|BasisFunction]]><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== blindkriging ===<br />
Build blind kriging models<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#AdaptiveModelBuilder|AdaptiveModelBuilder]]" combineOutputs="false"><br />
<Option key="nBestModels" value="1"/><br />
<br />
<!-- See the documentation for possible regression and correlation functions --><br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#KrigingFactory|KrigingFactory]]"><br />
<Option key="regressionMetric" value="cvpe"/><br />
<Option key="regressionFunction" value="regpoly0"/><br />
<Option key="multipleBasisFunctionsAllowed" value="false"/><br />
<br />
<Option key="initialHp" value="0.5"/><br />
<[[Config:BasisFunction|BasisFunction]]>corrgauss</[[Config:BasisFunction|BasisFunction]]><br />
<br />
<[[Config:Optimizer|Optimizer]]>fminconWithDerivatives</[[Config:Optimizer|Optimizer]]><br />
</[[Config:ModelFactory|ModelFactory]]><br />
<br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== kriging ===<br />
Build kriging models using the maximum likelihood to set the thetas<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#AdaptiveModelBuilder|AdaptiveModelBuilder]]" combineOutputs="false"><br />
<Option key="nBestModels" value="1"/><br />
<br />
<!-- See the documentation for possible regression and correlation functions --><br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#KrigingFactory|KrigingFactory]]"><br />
<Option key="regressionMetric" value=""/><br />
<Option key="regressionFunction" value="regpoly0"/><br />
<Option key="multipleBasisFunctionsAllowed" value="false"/><br />
<br />
<Option key="initialHp" value="0.5"/><br />
<[[Config:BasisFunction|BasisFunction]]>corrgauss</[[Config:BasisFunction|BasisFunction]]><br />
<br />
<[[Config:Optimizer|Optimizer]]>fminconWithDerivatives</[[Config:Optimizer|Optimizer]]><br />
</[[Config:ModelFactory|ModelFactory]]><br />
<br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== splinesgenetic ===<br />
Build spline models using a GA<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#GeneticModelBuilder|GeneticModelBuilder]]" combineOutputs="false"><br />
<!-- Re-start strategy for resuming the optimization process between sampling iterations.<br />
One of 'random','continue','model' and 'intelligent' (Default). See the docs for more information --><br />
<Option key="restartStrategy" value="intelligent"/><br />
<Option key="populationType" value="doubleVector"/><br />
<Option key="populationSize" value="10"/><br />
<Option key="maxGenerations" value="10"/><br />
<Option key="eliteCount" value="1"/><br />
<Option key="crossoverFraction" value="0.7"/><br />
<Option key="stallGenLimit" value="4"/><br />
<Option key="stallTimeLimit" value="Inf"/><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#SplineFactory|SplineFactory]]"><br />
<Option key="creationFcn" value="@gacreationuniform"/><br />
<Option key="crossoverFcn" value="@crossoverheuristic"/><br />
<Option key="mutationFcn" value="@mutationadaptfeasible"/><br />
<!-- <Option key="creationFcn" value="createInitialPopulation"/><br />
<Option key="crossoverFcn" value="crossover"/><br />
<Option key="mutationFcn" value="mutation"/> --><br />
<br />
<Option key="smoothingBounds" value="0,1"/> <br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== splinessim ===<br />
Build spline models using the Simulated Annealing modelbuilder<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#OptimizerModelBuilder|OptimizerModelBuilder]]" combineOutputs="false"><br />
<!-- Re-start strategy for resuming the optimization process between sampling iterations.<br />
One of 'random','continue','model' and 'intelligent' (Default). See the docs for more information --><br />
<Option key="restartStrategy" value="intelligent"/><br />
<br />
<[[Config:Optimizer|Optimizer]] type="[[Optimizer#MatlabSimAnnealing|MatlabSimAnnealing]]"><br />
<Option key="maxIterations" value="100"/><br />
<Option key="maxFunEvals" value="20"/><br />
</[[Config:Optimizer|Optimizer]]><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#SplineFactory|SplineFactory]]"><br />
<Option key="smoothingBounds" value="0,1"/> <br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== splinesps ===<br />
Build spline models using the Pattern Search modelbuilder<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#OptimizerModelBuilder|OptimizerModelBuilder]]" combineOutputs="false"><br />
<!-- Re-start strategy for resuming the optimization process between sampling iterations.<br />
One of 'random','continue','model' and 'intelligent' (Default). See the docs for more information --><br />
<Option key="restartStrategy" value="intelligent"/><br />
<br />
<[[Config:Optimizer|Optimizer]] type="[[Optimizer#MatlabPatternSearch|MatlabPatternSearch]]"><br />
<Option key="maxIterations" value="100"/><br />
<Option key="maxFunEvals" value="20"/><br />
<Option key="searchMethod" value="GPSPositiveBasis2N"/><br />
<Option key="pollMethod" value="MADSPositiveBasis2N"/> <br />
</[[Config:Optimizer|Optimizer]]><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#SplineFactory|SplineFactory]]"><br />
<Option key="smoothingBounds" value="0,1"/> <br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== splinesoptim ===<br />
Use the Matlab optimization toolbox to build spline models<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#OptimizerModelBuilder|OptimizerModelBuilder]]" combineOutputs="false"><br />
<!-- Re-start strategy for resuming the optimization process between sampling iterations.<br />
One of 'random','continue','model' and 'intelligent' (Default). See the docs for more information --><br />
<Option key="restartStrategy" value="intelligent"/><br />
<br />
<[[Config:Optimizer|Optimizer]] type="[[Optimizer#MatlabOptimizer|MatlabOptimizer]]"><br />
<Option key="maxIterations" value="100"/><br />
<Option key="maxFunEvals" value="20"/><br />
</[[Config:Optimizer|Optimizer]]><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#SplineFactory|SplineFactory]]"><br />
<Option key="smoothingBounds" value="0,1"/> <br />
</[[Config:ModelFactory|ModelFactory]]><br />
<br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== ipol ===<br />
Simple linear/cubic/nearest neighbour interpolation models for scattered data. For one D uses interp1, for nD uses griddata(n)<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#AdaptiveModelBuilder|AdaptiveModelBuilder]]" combineOutputs="false"><br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#InterpolationFactory|InterpolationFactory]]"><br />
<!-- depending on the input dimension options are: linear, nearest, and cubic<br />
if you are using Matlab r2009 or later you can also use 'natural' --><br />
<Option key="method" value="linear"/> <br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== ann ===<br />
Use a custom evolutionary-like strategy to generate ANN models, this is much faster than the GA approach but not necessarily better<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#SequentialModelBuilder|SequentialModelBuilder]]" combineOutputs="false"><br />
<Option key="maximumRunLength" value="15"/><br />
<!-- Degeneration of score if a model gets older --><br />
<Option key="decay" value=".99"/><br />
<!-- Size of the best model history --><br />
<Option key="historySize" value="6"/><br />
<!-- One of best, last. When set to best the best `historySize' models are kept,<br />
- - when set to last, the last `historySize' models are kept --><br />
<Option key="strategy" value="best"/><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#ANNFactory|ANNFactory]]"><br />
<!--initial hidden layer dimension--><br />
<Option key="initialSize" value="3,3"/><br />
<!--comma separated list of allowed learning rules--><br />
<Option key="allowedLearningRules" value="trainbr"/><br />
<!--performance function to use, empty means use the matlab default--><br />
<Option key="performFcn" value=""/><br />
<!--how many epochs to train for--><br />
<Option key="epochs" value="300"/><br />
<!--max time to train for--><br />
<Option key="trainingTime" value="Inf"/><br />
<!--range of initial random weights--><br />
<Option key="initWeightRange" value="-0.8,0.8"/><br />
<!--mutation changes x neurons at a time (in a random layer) with x in [lb ub]--><br />
<Option key="hiddenUnitDelta" value="-2,3"/><br />
<!--train until the error reaches this goal--><br />
<Option key="trainingGoal" value="0"/><br />
<!--show training progress every x epochs, set to NaN to disable--><br />
<Option key="trainingProgress" value="NaN"/><br />
<!--How to train the network, one of 'auto' or 'earlyStopping'<br />
auto: train with early stopping unless regularization is employed<br />
Set to any other value for simply training on all the data, doing nothing special --><br />
<Option key="trainMethod" value="auto"/><br />
<!--the training set - validation set - testset ratios--><br />
<Option key="earlyStoppingRatios" value="0.80,0.20,0"/><br />
<!-- Transfer function to use for all hidden layers and the output layer<br />
So should be a list of max 2 items --><br />
<Option key="transferFunctionTemplate" value="tansig,purelin"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== anngenetic ===<br />
Use the matlab gads toolbox to select ANN parameters using a GA<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#GeneticModelBuilder|GeneticModelBuilder]]" combineOutputs="false"><br />
<!-- Re-start strategy for resuming the optimization process between sampling iterations.<br />
One of 'random','continue','model' and 'intelligent' (Default). See the docs for more information --><br />
<Option key="restartStrategy" value="continue"/><br />
<!--See that matlab gads toolbox documentation for more information on the options--><br />
<Option key="populationType" value="custom"/><br />
<Option key="populationSize" value="10"/><br />
<Option key="crossoverFraction" value="0.7"/><br />
<Option key="maxGenerations" value="10"/><br />
<Option key="eliteCount" value="1"/><br />
<Option key="stallGenLimit" value="4"/><br />
<Option key="stallTimeLimit" value="Inf"/><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#ANNFactory|ANNFactory]]"><br />
<Option key="crossoverFcn" value="crossover"/><br />
<Option key="mutationFcn" value="mutation"/><br />
<Option key="creationFcn" value="createInitialPopulation"/><br />
<br />
<!--initial hidden layer dimension--><br />
<Option key="initialSize" value="3,3"/><br />
<!--comma separated list of allowed learning rules--><br />
<Option key="allowedLearningRules" value="trainbr"/><br />
<!--performance function to use, empty means use the matlab default--><br />
<Option key="performFcn" value=""/><br />
<!--how many epochs to train for--><br />
<Option key="epochs" value="300"/><br />
<!--max time to train for--><br />
<Option key="trainingTime" value="Inf"/><br />
<!--range of initial random weights--><br />
<Option key="initWeightRange" value="-0.8,0.8"/><br />
<!--mutation changes x neurons at a time (in a random layer) with x in [lb ub]--><br />
<Option key="hiddenUnitDelta" value="-2,3"/><br />
<!--train until the error reaches this goal--><br />
<Option key="trainingGoal" value="0"/><br />
<!--show training progress every x epochs, set to NaN to disable--><br />
<Option key="trainingProgress" value="NaN"/><br />
<!--How to train the network, one of 'auto' or 'earlyStopping'<br />
auto: train with early stopping unless regularization is employed<br />
Set to any other value for simply training on all the data, doing nothing special --><br />
<Option key="trainMethod" value="auto"/><br />
<!--the training set - validation set - testset ratios--><br />
<Option key="earlyStoppingRatios" value="0.80,0.20,0"/><br />
<!-- Transfer function to use for all hidden layers and the output layer<br />
So should be a list of max 2 items --><br />
<Option key="transferFunctionTemplate" value="tansig,purelin"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== annfixed ===<br />
Fixed ANN model builder, allows you to choose the hidden layer structure manually Thus there is no optimization algorithm involved.<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#AdaptiveModelBuilder|AdaptiveModelBuilder]]" combineOutputs="false"><br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#ANNFactory|ANNFactory]]"><br />
<Option key="allowedLearningRules" value="trainbr"/><br />
<Option key="initialSize" value="3,3"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== annrandom ===<br />
Random ANN model builder, usefull as a baseline comparison<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#RandomModelBuilder|RandomModelBuilder]]" combineOutputs="false"><br />
<!--This many iterations before allowing new samples--><br />
<Option key="runSize" value="10"/><br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#ANNFactory|ANNFactory]]"><br />
<Option key="allowedLearningRules" value="trainbr,trainlm,trainscg"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== fanngenetic ===<br />
Use the matlab gads toolbox to select ANN parameters using a GA (based on the FANN library)<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#GeneticModelBuilder|GeneticModelBuilder]]" combineOutputs="false"><br />
<!-- Re-start strategy for resuming the optimization process between sampling iterations.<br />
One of 'random','continue','model' and 'intelligent' (Default). See the docs for more information --><br />
<Option key="restartStrategy" value="continue"/><br />
<!--See that matlab gads toolbox documentation for more information on the options--> <br />
<Option key="populationType" value="custom"/> <br />
<Option key="populationSize" value="10"/> <br />
<Option key="crossoverFraction" value="0.7"/> <br />
<Option key="maxGenerations" value="10"/> <br />
<Option key="eliteCount" value="1"/> <br />
<Option key="stallGenLimit" value="4"/> <br />
<Option key="stallTimeLimit" value="Inf"/> <br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#FANNFactory|FANNFactory]]"> <br />
<Option key="crossoverFcn" value="crossover"/> <br />
<Option key="mutationFcn" value="mutation"/> <br />
<Option key="creationFcn" value="createInitialPopulation"/> <br />
<br />
<!--initial hidden layer dimension--> <br />
<Option key="initialSize" value="4,4"/> <br />
<!--how many epochs to train for--> <br />
<Option key="epochs" value="1500"/> <br />
<!--range of initial random weights--> <br />
<Option key="initWeightRange" value="-0.8,0.8"/> <br />
<!--mutation changes x neurons at a time (in a random layer) with x in [lb ub]--> <br />
<Option key="hiddenUnitDelta" value="-2,2"/> <br />
<!--train until the error reaches this goal--> <br />
<Option key="trainingGoal" value="0"/> <br />
</[[Config:ModelFactory|ModelFactory]]> <br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== nanngenetic ===<br />
Use the matlab gads toolbox to select ANN parameters using a GA (based on the NNSYSID library)<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#GeneticModelBuilder|GeneticModelBuilder]]" combineOutputs="false"><br />
<!-- Re-start strategy for resuming the optimization process between sampling iterations.<br />
One of 'random','continue','model' and 'intelligent' (Default). See the docs for more information --><br />
<Option key="restartStrategy" value="continue"/><br />
<!--See that matlab gads toolbox documentation for more information on the options--> <br />
<Option key="populationType" value="custom"/> <br />
<Option key="populationSize" value="10"/> <br />
<Option key="crossoverFraction" value="0.7"/> <br />
<Option key="maxGenerations" value="10"/> <br />
<Option key="eliteCount" value="1"/> <br />
<Option key="stallGenLimit" value="4"/> <br />
<Option key="stallTimeLimit" value="Inf"/> <br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#NANNFactory|NANNFactory]]"> <br />
<Option key="crossoverFcn" value="crossover"/> <br />
<Option key="mutationFcn" value="mutation"/> <br />
<Option key="creationFcn" value="createInitialPopulation"/> <br />
<br />
<!--initial hidden layer dimension--> <br />
<Option key="initialSize" value="10"/> <br />
<!--how many epochs to train for--> <br />
<Option key="epochs" value="500"/> <br />
<!--range of initial random weights--> <br />
<Option key="initWeightRange" value="-0.8,0.8"/> <br />
<!--mutation changes x neurons at a time (in a random layer) with x in [lb ub]--> <br />
<Option key="hiddenUnitDelta" value="-2,3"/> <br />
<!-- pruning techniques used : 0: none, 1: Mag Threshold, 2: Iterative Mag, 3: OBD, 4: OBS --> <br />
<Option key="allowedPruneTechniques" value="0,1,2,3,4"/><br />
<!-- threshold for magnitude based pruning --> <br />
<Option key="threshold" value="0.2"/> <br />
<!-- retrain epochs while pruning--> <br />
<Option key="retrain" value="50"/> <br />
</[[Config:ModelFactory|ModelFactory]]> <br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== lssvmgenetic ===<br />
Use the matlab gads toolbox to select LSSVM parameters using a GA (based on LSSVM-lab)<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#GeneticModelBuilder|GeneticModelBuilder]]" combineOutputs="false"><br />
<!-- Re-start strategy for resuming the optimization process between sampling iterations.<br />
One of 'random','continue','model' and 'intelligent' (Default). See the docs for more information --><br />
<Option key="restartStrategy" value="intelligent"/><br />
<!-- Plot the optimization surface, visualizes the search through the parameter space (2D only) --><br />
<Option key="plotOptimSurface" value="false"/><br />
<Option key="populationType" value="doubleVector"/><br />
<Option key="populationSize" value="10"/><br />
<Option key="maxGenerations" value="10"/><br />
<Option key="eliteCount" value="1"/><br />
<Option key="crossoverFraction" value="0.7"/><br />
<Option key="stallGenLimit" value="4"/><br />
<Option key="stallTimeLimit" value="Inf"/><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#SVMFactory|SVMFactory]]"><br />
<Option key="creationFcn" value="@gacreationuniform"/><br />
<Option key="crossoverFcn" value="@crossoverheuristic"/><br />
<Option key="mutationFcn" value="@mutationadaptfeasible"/><br />
<!-- <Option key="creationFcn" value="createInitialPopulation"/><br />
<Option key="crossoverFcn" value="crossover"/><br />
<Option key="mutationFcn" value="mutation"/> --><br />
<br />
<Option key="backend" value="lssvm"/><br />
<Option key="kernel" value="rbf"/><br />
<Option key="kernelParamBounds" value="-2,2"/><br />
<Option key="regParamBounds" value="-5,5"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== lssvmps ===<br />
Use the matlab gads toolbox to select LSSVM parameters using Pattern Search<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#OptimizerModelBuilder|OptimizerModelBuilder]]" combineOutputs="false"><br />
<!-- Re-start strategy for resuming the optimization process between sampling iterations.<br />
One of 'random','continue','model' and 'intelligent' (Default). See the docs for more information --><br />
<Option key="restartStrategy" value="intelligent"/><br />
<!-- Plot the optimization surface, visualizes the search through the parameter space (2D only) --><br />
<Option key="plotOptimSurface" value="false"/><br />
<br />
<!--See that matlab gads toolbox documentation for more information on the options--> <br />
<[[Config:Optimizer|Optimizer]] type="[[Optimizer#MatlabPatternSearch|MatlabPatternSearch]]"><br />
<Option key="maxIterations" value="100"/><br />
<Option key="maxFunEvals" value="20"/><br />
<Option key="searchMethod" value="GPSPositiveBasis2N"/><br />
<Option key="pollMethod" value="MADSPositiveBasis2N"/> <br />
</[[Config:Optimizer|Optimizer]]><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#SVMFactory|SVMFactory]]"><br />
<Option key="backend" value="lssvm"/><br />
<Option key="kernel" value="rbf"/><br />
<Option key="kernelParamBounds" value="-2,2"/><br />
<Option key="regParamBounds" value="-5,5"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== lssvmoptim ===<br />
Use the matlab optimization toolbox to select LSSVM parameters<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#OptimizerModelBuilder|OptimizerModelBuilder]]" combineOutputs="false"><br />
<!-- Re-start strategy for resuming the optimization process between sampling iterations.<br />
One of 'random','continue','model' and 'intelligent' (Default). See the docs for more information --><br />
<Option key="restartStrategy" value="intelligent"/><br />
<!-- Plot the optimization surface, visualizes the search through the parameter space (2D only) --><br />
<Option key="plotOptimSurface" value="false"/><br />
<!--See the interface matlab file and the optimization toolbox documentation for more information on the options--><br />
<[[Config:Optimizer|Optimizer]] type="[[Optimizer#MatlabOptimizer|MatlabOptimizer]]"><br />
<Option key="maxIterations" value="100"/><br />
<Option key="maxFunEvals" value="20"/><br />
</[[Config:Optimizer|Optimizer]]><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#SVMFactory|SVMFactory]]"><br />
<Option key="backend" value="lssvm"/><br />
<Option key="kernel" value="rbf"/><br />
<Option key="kernelParamBounds" value="-2,2"/><br />
<Option key="regParamBounds" value="-5,5"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== lssvmpso ===<br />
Use the PSO toolbox to select LSSVM parameters using Particle Swarm Optimization<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#OptimizerModelBuilder|OptimizerModelBuilder]]" combineOutputs="false"><br />
<!-- Re-start strategy for resuming the optimization process between sampling iterations.<br />
One of 'random','continue','model' and 'intelligent' (Default). See the docs for more information --><br />
<Option key="restartStrategy" value="intelligent"/><br />
<!-- Plot the optimization surface, visualizes the search through the parameter space (2D only) --><br />
<Option key="plotOptimSurface" value="false"/><br />
<br />
<[[Config:Optimizer|Optimizer]] type="[[Optimizer#PSOtOptimizer|PSOtOptimizer]]"><br />
<Option key="typePSO" value="0"/><br />
<Option key="seedPSO" value="1"/><br />
<Option key="popSize" value="10"/><br />
<Option key="maxiters" value="10"/><br />
<Option key="epochInertia" value="8"/><br />
<Option key="gradientTermination" value="8"/><br />
</[[Config:Optimizer|Optimizer]]><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#SVMFactory|SVMFactory]]"><br />
<Option key="backend" value="lssvm"/><br />
<Option key="kernel" value="rbf"/><br />
<Option key="kernelParamBounds" value="-2,2"/><br />
<Option key="regParamBounds" value="-5,5"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== lssvmsim ===<br />
Use the matlab gads toolbox to select LSSVM parameters using simulated annealing<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#OptimizerModelBuilder|OptimizerModelBuilder]]" combineOutputs="false"><br />
<!-- Re-start strategy for resuming the optimization process between sampling iterations.<br />
One of 'random','continue','model' and 'intelligent' (Default). See the docs for more information --><br />
<Option key="restartStrategy" value="intelligent"/><br />
<!-- Plot the optimization surface, visualizes the search through the parameter space (2D only) --><br />
<Option key="plotOptimSurface" value="false"/><br />
<!--See the interface matlab file and the gads toolbox documentation for more information on the options--><br />
<[[Config:Optimizer|Optimizer]] type="[[Optimizer#MatlabSimAnnealing|MatlabSimAnnealing]]"><br />
<Option key="maxIterations" value="100"/><br />
<Option key="maxFunEvals" value="20"/><br />
</[[Config:Optimizer|Optimizer]]><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#SVMFactory|SVMFactory]]"><br />
<Option key="backend" value="lssvm"/><br />
<Option key="kernel" value="rbf"/><br />
<Option key="kernelParamBounds" value="-2,2"/><br />
<Option key="regParamBounds" value="-5,5"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== lssvmdirect ===<br />
Use the DIviding RECtangles algorithm to optimize the LS-SVM hyperparameters<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#OptimizerModelBuilder|OptimizerModelBuilder]]" combineOutputs="false"><br />
<!-- Re-start strategy for resuming the optimization process between sampling iterations.<br />
One of 'random','continue','model' and 'intelligent' (Default). See the docs for more information --><br />
<Option key="restartStrategy" value="intelligent"/><br />
<!-- Plot the optimization surface, visualizes the search through the parameter space (2D only) --><br />
<Option key="plotOptimSurface" value="false"/><br />
<br />
<[[Config:Optimizer|Optimizer]] type="[[Optimizer#DirectOptimizer|DirectOptimizer]]"><br />
<Option key="maxits" value="100"/><br />
<Option key="maxevals" value="20"/><br />
</[[Config:Optimizer|Optimizer]]><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#SVMFactory|SVMFactory]]"><br />
<Option key="backend" value="lssvm"/><br />
<Option key="kernel" value="rbf"/><br />
<Option key="kernelParamBounds" value="-2,2"/><br />
<Option key="regParamBounds" value="-5,5"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== lssvmego ===<br />
Generate LS-SVM models using Efficient Global Optimization (EGO). This means internally a kriging model is constructed to predict where one can expect to find good model parameters<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#EGOModelBuilder|EGOModelBuilder]]" combineOutputs="false"><br />
<Option key="numIterations" value="10"/><br />
<Option key="initPopSize" value="5"/><br />
<Option key="restartStrategy" value="continue"/><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#SVMFactory|SVMFactory]]"><br />
<Option key="backend" value="lssvm"/><br />
<Option key="kernel" value="rbf"/><br />
<Option key="kernelParamBounds" value="-4,4"/><br />
<Option key="regParamBounds" value="-5,5"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
<br />
<!-- Optimizer for the internal kriging model --><br />
<[[Config:Optimizer|Optimizer]]>fminconWithDerivatives</[[Config:Optimizer|Optimizer]]><br />
<br />
<!-- Sampleselector to use --><br />
<[[Config:SampleSelector|SampleSelector]]>expectedImprovement</[[Config:SampleSelector|SampleSelector]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== lssvmrandom ===<br />
Generate random LSSVM models, useful as a baseline comparison<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#RandomModelBuilder|RandomModelBuilder]]" combineOutputs="false"><br />
<!-- Plot the optimization surface, visualizes the search through the parameter space (2D only) --><br />
<Option key="plotOptimSurface" value="false"/><br />
<br />
<Option key="runSize" value="20"/><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#SVMFactory|SVMFactory]]"><br />
<Option key="backend" value="lssvm"/><br />
<Option key="kernel" value="rbf"/><br />
<Option key="kernelParamBounds" value="-2,2"/><br />
<Option key="regParamBounds" value="-5,5"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== svmgenetic ===<br />
Use the matlab gads toolbox to select SVM parameters using a GA (based on libsvm)<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#GeneticModelBuilder|GeneticModelBuilder]]" combineOutputs="false"><br />
<!-- Re-start strategy for resuming the optimization process between sampling iterations.<br />
One of 'random','continue','model' and 'intelligent' (Default). See the docs for more information --><br />
<Option key="restartStrategy" value="intelligent"/><br />
<!-- Plot the optimization surface, visualizes the search through the parameter space (2D only) --><br />
<Option key="plotOptimSurface" value="false"/><br />
<!--See the interface matlab file and the gads toolbox documentation for more information on the options--><br />
<Option key="populationType" value="doubleVector"/><br />
<Option key="populationSize" value="10"/><br />
<Option key="maxGenerations" value="10"/><br />
<Option key="eliteCount" value="1"/><br />
<Option key="crossoverFraction" value="0.7"/><br />
<Option key="stallGenLimit" value="4"/><br />
<Option key="stallTimeLimit" value="Inf"/><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#SVMFactory|SVMFactory]]"><br />
<Option key="creationFcn" value="@gacreationuniform"/><br />
<Option key="crossoverFcn" value="@crossoverheuristic"/><br />
<Option key="mutationFcn" value="@mutationadaptfeasible"/><br />
<!-- <Option key="creationFcn" value="createInitialPopulation"/><br />
<Option key="crossoverFcn" value="crossover"/><br />
<Option key="mutationFcn" value="mutation"/>--><br />
<br />
<Option key="backend" value="libSVM"/><br />
<Option key="type" value="epsilon-SVR"/><br />
<Option key="kernel" value="rbf"/><br />
<Option key="kernelParamBounds" value="-4,4"/><br />
<Option key="regParamBounds" value="-5,5"/><br />
<Option key="nu" value="0.01"/><br />
<Option key="epsilon" value="0"/><br />
<Option key="stoppingTolerance" value="1e-6"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== svmps ===<br />
Use the matlab gads toolbox to select SVM parameters using Pattern Search<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#OptimizerModelBuilder|OptimizerModelBuilder]]" combineOutputs="false"><br />
<!-- Re-start strategy for resuming the optimization process between sampling iterations.<br />
One of 'random','continue','model' and 'intelligent' (Default). See the docs for more information --><br />
<Option key="restartStrategy" value="intelligent"/><br />
<!-- Plot the optimization surface, visualizes the search through the parameter space (2D only) --><br />
<Option key="plotOptimSurface" value="false"/><br />
<br />
<!--See the interface matlab file and the gads toolbox documentation for more information on the options--><br />
<[[Config:Optimizer|Optimizer]] type="[[Optimizer#MatlabPatternSearch|MatlabPatternSearch]]"><br />
<Option key="maxIterations" value="100"/><br />
<Option key="maxFunEvals" value="20"/><br />
<Option key="searchMethod" value="GPSPositiveBasis2N"/><br />
<Option key="pollMethod" value="MADSPositiveBasis2N"/> <br />
</[[Config:Optimizer|Optimizer]]><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#SVMFactory|SVMFactory]]"><br />
<Option key="backend" value="libSVM"/><br />
<Option key="type" value="epsilon-SVR"/><br />
<Option key="kernel" value="rbf"/><br />
<Option key="kernelParamBounds" value="-2,2"/><br />
<Option key="regParamBounds" value="-5,5"/><br />
<Option key="nu" value="0.01"/><br />
<Option key="epsilon" value="0"/><br />
<Option key="stoppingTolerance" value="1e-6"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== svmsim ===<br />
Use the matlab gads toolbox to select SVM parameters using simulated annealing<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#OptimizerModelBuilder|OptimizerModelBuilder]]" combineOutputs="false"><br />
<!-- Re-start strategy for resuming the optimization process between sampling iterations.<br />
One of 'random','continue','model' and 'intelligent' (Default). See the docs for more information --><br />
<Option key="restartStrategy" value="intelligent"/><br />
<!-- Plot the optimization surface, visualizes the search through the parameter space (2D only) --><br />
<Option key="plotOptimSurface" value="false"/><br />
<!--See the interface matlab file and the gads toolbox documentation for more information on the options--> <br />
<[[Config:Optimizer|Optimizer]] type="[[Optimizer#MatlabSimAnnealing|MatlabSimAnnealing]]"><br />
<Option key="maxIterations" value="100"/><br />
<Option key="maxFunEvals" value="20"/><br />
</[[Config:Optimizer|Optimizer]]><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#SVMFactory|SVMFactory]]"><br />
<Option key="backend" value="libSVM"/><br />
<Option key="type" value="epsilon-SVR"/><br />
<Option key="kernel" value="rbf"/><br />
<Option key="kernelParamBounds" value="-2,2"/><br />
<Option key="regParamBounds" value="-5,5"/><br />
<Option key="nu" value="0.01"/><br />
<Option key="epsilon" value="0"/><br />
<Option key="stoppingTolerance" value="1e-6"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== svmoptim ===<br />
Use the matlab optimization toolbox to select SVM parameters<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#OptimizerModelBuilder|OptimizerModelBuilder]]" combineOutputs="false"><br />
<!-- Re-start strategy for resuming the optimization process between sampling iterations.<br />
One of 'random','continue','model' and 'intelligent' (Default). See the docs for more information --><br />
<Option key="restartStrategy" value="intelligent"/><br />
<!-- Plot the optimization surface, visualizes the search through the parameter space (2D only) --><br />
<Option key="plotOptimSurface" value="false"/><br />
<!--See the interface matlab file and the optimization toolbox documentation for more<br />
information on the options--><br />
<[[Config:Optimizer|Optimizer]] type="[[Optimizer#MatlabOptimizer|MatlabOptimizer]]"><br />
<Option key="maxIterations" value="100"/><br />
<Option key="maxFunEvals" value="20"/><br />
</[[Config:Optimizer|Optimizer]]><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#SVMFactory|SVMFactory]]"><br />
<Option key="backend" value="libSVM"/><br />
<Option key="type" value="epsilon-SVR"/><br />
<Option key="kernel" value="rbf"/><br />
<Option key="kernelParamBounds" value="-2,2"/><br />
<Option key="regParamBounds" value="-5,5"/><br />
<Option key="nu" value="0.01"/><br />
<Option key="epsilon" value="0"/><br />
<Option key="stoppingTolerance" value="1e-6"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== svmpso ===<br />
Use the PSO toolbox to select SVM parameters using Particle Swarm Optimization<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#OptimizerModelBuilder|OptimizerModelBuilder]]" combineOutputs="false"><br />
<!-- Re-start strategy for resuming the optimization process between sampling iterations.<br />
One of 'random','continue','model' and 'intelligent' (Default). See the docs for more information --><br />
<Option key="restartStrategy" value="intelligent"/><br />
<!-- Plot the optimization surface, visualizes the search through the parameter space (2D only) --><br />
<Option key="plotOptimSurface" value="false"/><br />
<br />
<[[Config:Optimizer|Optimizer]] type="[[Optimizer#PSOtOptimizer|PSOtOptimizer]]"><br />
<Option key="typePSO" value="0"/><br />
<Option key="seedPSO" value="1"/><br />
<Option key="popSize" value="10"/><br />
<Option key="maxiters" value="10"/><br />
<Option key="epochInertia" value="8"/><br />
<Option key="gradientTermination" value="8"/><br />
</[[Config:Optimizer|Optimizer]]><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#SVMFactory|SVMFactory]]"><br />
<Option key="backend" value="libSVM"/><br />
<Option key="type" value="epsilon-SVR"/><br />
<Option key="kernel" value="rbf"/><br />
<Option key="kernelParamBounds" value="-2,2"/><br />
<Option key="regParamBounds" value="-5,5"/><br />
<Option key="nu" value="0.01"/><br />
<Option key="epsilon" value="0"/><br />
<Option key="stoppingTolerance" value="1e-6"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== svmdirect ===<br />
Use the DIviding RECtangles algorithm to optimize the SVM hyperparameters<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#OptimizerModelBuilder|OptimizerModelBuilder]]" combineOutputs="false"><br />
<!-- Re-start strategy for resuming the optimization process between sampling iterations.<br />
One of 'random','continue','model' and 'intelligent' (Default). See the docs for more information --><br />
<Option key="restartStrategy" value="intelligent"/><br />
<!-- Plot the optimization surface, visualizes the search through the parameter space (2D only) --><br />
<Option key="plotOptimSurface" value="false"/><br />
<br />
<[[Config:Optimizer|Optimizer]] type="[[Optimizer#DirectOptimizer|DirectOptimizer]]"><br />
<Option key="maxits" value="100"/><br />
<Option key="maxevals" value="20"/><br />
</[[Config:Optimizer|Optimizer]]><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#SVMFactory|SVMFactory]]"><br />
<Option key="backend" value="libSVM"/><br />
<Option key="type" value="epsilon-SVR"/><br />
<Option key="kernel" value="rbf"/><br />
<Option key="kernelParamBounds" value="-2,2"/><br />
<Option key="regParamBounds" value="-5,5"/><br />
<Option key="nu" value="0.01"/><br />
<Option key="epsilon" value="0"/><br />
<Option key="stoppingTolerance" value="1e-6"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== svmrandom ===<br />
Generate random SVMs, useful as a baseline comparison<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#RandomModelBuilder|RandomModelBuilder]]" combineOutputs="false"><br />
<!-- Plot the optimization surface, visualizes the search through the parameter space (2D only) --><br />
<Option key="plotOptimSurface" value="false"/><br />
<br />
<Option key="runSize" value="20"/><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#SVMFactory|SVMFactory]]"><br />
<Option key="backend" value="libSVM"/><br />
<Option key="type" value="epsilon-SVR"/><br />
<Option key="kernel" value="rbf"/><br />
<Option key="kernelParamBounds" value="-2,2"/><br />
<Option key="regParamBounds" value="-5,5"/><br />
<Option key="nu" value="0.01"/><br />
<Option key="epsilon" value="0"/><br />
<Option key="stoppingTolerance" value="1e-6"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== rbfnngenetic ===<br />
Genetic model builder for Radial Basis Function Neural networks<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#GeneticModelBuilder|GeneticModelBuilder]]" combineOutputs="false"><br />
<!-- Re-start strategy for resuming the optimization process between sampling iterations.<br />
One of 'random','continue','model' and 'intelligent' (Default). See the docs for more information --><br />
<Option key="restartStrategy" value="intelligent"/><br />
<Option key="populationType" value="doubleVector"/><br />
<Option key="populationSize" value="10"/><br />
<Option key="maxGenerations" value="10"/><br />
<Option key="eliteCount" value="1"/><br />
<Option key="crossoverFraction" value="0.7"/><br />
<Option key="stallGenLimit" value="4"/><br />
<Option key="stallTimeLimit" value="Inf"/><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#RBFNNFactory|RBFNNFactory]]"><br />
<Option key="creationFcn" value="@gacreationuniform"/><br />
<Option key="crossoverFcn" value="@crossoverheuristic"/><br />
<Option key="mutationFcn" value="@mutationadaptfeasible"/><br />
<!-- <Option key="creationFcn" value="createInitialPopulation"/><br />
<Option key="crossoverFcn" value="crossover"/><br />
<Option key="mutationFcn" value="mutation"/> --><br />
<br />
<!--Error goal when constructing the network--><br />
<Option key="goal" value="0"/><br />
<!--Initial value for the spread --><br />
<Option key="spread" value="1"/><br />
<!--Spread bounds --><br />
<Option key="spreadBounds" value="0.0001,30"/><br />
<!--Maximum number of neurons to use per network--><br />
<Option key="maxNeurons" value="500"/><br />
<Option key="trainingProgress" value="Inf"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== rbfnnps ===<br />
Build Radial Basis Function Neural networks using Pattern Search<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#OptimizerModelBuilder|OptimizerModelBuilder]]" combineOutputs="false"><br />
<!-- Re-start strategy for resuming the optimization process between sampling iterations.<br />
One of 'random','continue','model' and 'intelligent' (Default). See the docs for more information --><br />
<Option key="restartStrategy" value="intelligent"/><br />
<br />
<[[Config:Optimizer|Optimizer]] type="[[Optimizer#MatlabPatternSearch|MatlabPatternSearch]]"><br />
<Option key="maxIterations" value="100"/><br />
<Option key="maxFunEvals" value="20"/><br />
<Option key="searchMethod" value="GPSPositiveBasis2N"/><br />
<Option key="pollMethod" value="MADSPositiveBasis2N"/> <br />
</[[Config:Optimizer|Optimizer]]><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#RBFNNFactory|RBFNNFactory]]"><br />
<!--Error goal when constructing the network--><br />
<Option key="goal" value="0"/><br />
<!--Initial value for the spread --><br />
<Option key="spread" value="1"/><br />
<!--Spread bounds --><br />
<Option key="spreadBounds" value="0.0001,30"/><br />
<!--Maximum number of neurons to use per network--><br />
<Option key="maxNeurons" value="500"/><br />
<Option key="trainingProgress" value="Inf"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== rbfnnsim ===<br />
Build Radial Basis Function Neural networks using Simulated Annealing<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#OptimizerModelBuilder|OptimizerModelBuilder]]" combineOutputs="false"><br />
<!-- Re-start strategy for resuming the optimization process between sampling iterations.<br />
One of 'random','continue','model' and 'intelligent' (Default). See the docs for more information --><br />
<Option key="restartStrategy" value="intelligent"/><br />
<br />
<[[Config:Optimizer|Optimizer]] type="[[Optimizer#MatlabSimAnnealing|MatlabSimAnnealing]]"><br />
<Option key="maxIterations" value="100"/><br />
<Option key="maxFunEvals" value="20"/><br />
</[[Config:Optimizer|Optimizer]]><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#RBFNNFactory|RBFNNFactory]]"><br />
<!--Error goal when constructing the network--><br />
<Option key="goal" value="0"/><br />
<!--Initial value for the spread --><br />
<Option key="spread" value="1"/><br />
<!--Spread bounds --><br />
<Option key="spreadBounds" value="0.0001,30"/><br />
<!--Maximum number of neurons to use per network--><br />
<Option key="maxNeurons" value="500"/><br />
<Option key="trainingProgress" value="Inf"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source><br />
=== heterogenetic ===<br />
A heterogeneous genetic model builder. Uses a genetic algorithm with speciation (island model) to evolve different model types together. The models types compete against each other until the best model prevails. So this model builder is a way to automatically select the best model type.<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]] type="[[AdaptiveModelBuilder#GeneticModelBuilder|GeneticModelBuilder]]" combineOutputs="false"><br />
<Option key="restartStrategy" value="continue"/><br />
<Option key="populationType" value="custom"/><br />
<!-- the population size must match the number of model interfaces minus 1 --><br />
<Option key="populationSize" value="10,10,10"/><br />
<Option key="maxGenerations" value="10"/><br />
<Option key="crossoverFraction" value="0.7"/> <br />
<Option key="eliteCount" value="1"/><br />
<Option key="stallGenLimit" value="4"/><br />
<Option key="stallTimeLimit" value="Inf"/><br />
<Option key="migrationDirection" value="forward"/><br />
<Option key="migrationFraction" value="0.1"/><br />
<Option key="migrationInterval" value="4"/><br />
<!-- Do we want to prevent any model type going completely extinct --><br />
<Option key="extinctionPrevention" value="yes"/> <br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#HeterogeneousFactory|HeterogeneousFactory]]"><br />
<Option key="creationFcn" value="createInitialPopulation"/><br />
<Option key="crossoverFcn" value="crossover"/><br />
<Option key="mutationFcn" value="mutation"/><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#EnsembleFactory|EnsembleFactory]]"><br />
<Option key="crossoverFcn" value="crossover"/><br />
<Option key="mutationFcn" value="mutation"/><br />
<!-- the maximum ensemble size --><br />
<Option key="maxSize" value="4"/><br />
<!-- Ensemble members should differ this much percent --><br />
<Option key="equalityThreshold" value="0.05"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#SVMFactory|SVMFactory]]"><br />
<Option key="creationFcn" value="createInitialPopulation"/><br />
<Option key="crossoverFcn" value="crossover"/><br />
<Option key="mutationFcn" value="mutation"/><br />
<br />
<Option key="backend" value="lssvm"/><br />
<Option key="kernel" value="rbf"/><br />
<Option key="kernelParamBounds" value="-2,2"/><br />
<Option key="regParamBounds" value="-5,5"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#RationalFactory|RationalFactory]]"><br />
<Option key="crossoverFcn" value="crossover"/><br />
<Option key="mutationFcn" value="mutation"/><br />
<Option key="creationFcn" value="createInitialPopulation"/><br />
<Option key="weightBounds" value="1,40"/><br />
<Option key="percentBounds" value="1,100"/><br />
<Option key="percentRational" value="70"/><br />
<Option key="frequencyVariable" value="off"/><br />
<Option key="basis" value="power"/><br />
</[[Config:ModelFactory|ModelFactory]]><br />
<br />
<[[Config:ModelFactory|ModelFactory]] type="[[ModelFactory#BFFactory|BFFactory]]"><br />
<Option key="type" value="RBF"/><br />
<br />
<Option key="crossoverFcn" value="crossover"/><br />
<Option key="mutationFcn" value="mutation"/><br />
<Option key="creationFcn" value="createInitialPopulation"/><br />
<br />
<[[Config:BasisFunction|BasisFunction]] name="gaussian" min=".1" max="5" scale="ln"/><br />
<[[Config:BasisFunction|BasisFunction]] name="multiquadric" min=".1" max="5" scale="ln"/><br />
<[[Config:BasisFunction|BasisFunction]] name="exponential" min=".1,.5" max="5,2" scale="ln,lin"/><br />
<br />
<Option key="regression" value="-1,0,1,2"/><br />
<Option key="backend" value="Direct"/><br />
</[[Config:ModelFactory|ModelFactory]]> <br />
</[[Config:ModelFactory|ModelFactory]]><br />
</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
</source></div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=Config:ContextConfig&diff=5105Config:ContextConfig2010-03-25T10:36:54Z<p>Icouckuy: AutoConfig for SUMO 6.2</p>
<hr />
<div>'''Generated for SUMO toolbox version 7.0'''.<br />
''We are well aware that documentation is not always complete and possibly even out of date in some cases. We try to document everything as best we can but much is limited by available time and manpower. We are an university research group after all. The most up to date documentation can always be found (if not here) in the default.xml configuration file and, of course, in the source files. If something is unclear please dont hesitate to [[Reporting problems|ask]].''<br />
== ContextConfig ==<br />
=== OutputDirectory ===<br />
Specifies the directory to use to store the results, logs, ... An absolute path can be specified, if not it is relative to the project directory.<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<br />
<!--Specifies the directory to use to store the results, logs, ... An absolute path can be specified, if not it is relative to the project directory.--><br />
<[[Config:OutputDirectory|OutputDirectory]]>output</[[Config:OutputDirectory|OutputDirectory]]></source><br />
=== PlotOptions ===<br />
Several options for configuring the intermediate plots can be found here<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<br />
<!--Several options for configuring the intermediate plots can be found here--><br />
<[[Config:PlotOptions|PlotOptions]]><br />
<br />
<!-- save model to disk as Matlab .mat file --><br />
<Option key="saveModels" value="true"/><br />
<br />
<!-- plot models model plots to disk --><br />
<Option key="plotModels" value="true"/><br />
<br />
<!-- contours are only available for 2 dimensions --><br />
<br />
<!-- also show the contour lines underneath a surface plot --><br />
<Option key="withContour" value="true"/><br />
<!-- use a contour plot instead of a surface plot --><br />
<Option key="plotContour" value="false"/><br />
<!-- also plot the model uncertainty, only supported by some models types --><br />
<Option key="plotUncertainty" value="false"/><br />
<br />
<!--<br />
Other options to customize plotting:<br />
<br />
<Option key="plotPoints" value="true"/><br />
<Option key="lighting" value="false"/><br />
<Option key="slices" value="3"/><br />
<Option key="grayScale" value="false"/><br />
<Option key="meshSize" value="41"/><br />
<Option key="fontSize" value="14"/><br />
<Option key="logScale" value="false"/><br />
--><br />
<br />
<!-- Output file format of the plot: png (default),eps,bmp,jpg,... --><br />
<Option key="outputType" value="png"/><br />
</[[Config:PlotOptions|PlotOptions]]></source><br />
=== Custom Options ===<br />
Available options:<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<br />
<!--Keep models from previous model iterations. See: http://www.sumowiki.intec.ugent.be/index.php/FAQ#I_sometimes_see_the_error_of_the_best_model_go_up.2C_shouldn.27t_it_decrease_monotonically.3F--><br />
<Option key="keepOldModels" value="off"/><br />
<!--If set to on the parallelMode option will check if the Matlab Parallel Computing Toolbox is installed. If so, model construction will occur in parallel where possible, thus speeding things up. Note that this option is still somewhat experimental. If you get weird problems switch it off again.--><br />
<Option key="parallelMode" value="off"/></source><br />
=== Profiling ===<br />
Configuration options for the profilers. Profilers allow you to track the modeling process and understand what happens. Within a profiler block, one or more &lt;Output type="type"/&gt; specifiers should be present. Possible types are toFile, toPanel, toTable, and toImage.<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<br />
<!--Configuration options for the profilers. Profilers allow you to track the modeling process and understand what happens. Within a profiler block, one or more <[[Config:Output|Output]] type="[[Output#type|type]]"/> specifiers should be present. Possible types are toFile, toPanel, toTable, and toImage.--><br />
<[[Config:Profiling|Profiling]]><br />
<!--<br />
You can restrict the available profilers to use by specifying a wildcard to disable the docked output if you dont want a window containing all the profilers.<br />
To select all profilers, simply put ".*"<br />
--><br />
<[[Config:Profiler|Profiler]] name=".*SampleMinimum.*|.*Measure.*|.*BestModel.*|.*ElapsedTime.*|.*MemoryUse.*" enabled="true"><br />
<!-- note that the toImage/toPanel handlers are quite expensive to use, removing them will speed things up --><br />
<[[Config:Output|Output]] type="[[Output#toPanel|toPanel]]"/><br />
<[[Config:Output|Output]] type="[[Output#toImage|toImage]]"/><br />
<[[Config:Output|Output]] type="[[Output#toTable|toTable]]"/><br />
<[[Config:Output|Output]] type="[[Output#toFile|toFile]]"/><br />
</[[Config:Profiler|Profiler]]> <br />
</[[Config:Profiling|Profiling]]></source></div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=Config:LevelPlot&diff=5104Config:LevelPlot2010-03-25T10:36:25Z<p>Icouckuy: AutoConfig for SUMO 6.2</p>
<hr />
<div>'''Generated for SUMO toolbox version 7.0'''.<br />
''We are well aware that documentation is not always complete and possibly even out of date in some cases. We try to document everything as best we can but much is limited by available time and manpower. We are an university research group after all. The most up to date documentation can always be found (if not here) in the default.xml configuration file and, of course, in the source files. If something is unclear please dont hesitate to [[Reporting problems|ask]].''<br />
== LevelPlot ==<br />
=== Custom Options ===<br />
Available options:<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<br />
<!--This option enables the levelPlotProfiler (be sure to enable it in the profilers section too). A sample evaluator should provide the reference data set over which the error should be calculated.--><br />
<Option key="makeLevelPlots" value="off"/></source><br />
=== SampleEvaluator ===<br />
This tag defines the SampleEvaluator that is used by the levelplot (it needs to get its data from somewhere)<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<br />
<!--This tag defines the SampleEvaluator that is used by the levelplot (it needs to get its data from somewhere)--><br />
<[[Config:SampleEvaluator|SampleEvaluator]] type="ibbt.sumo.sampleevaluators.datasets.ScatteredDatasetSampleEvaluator"><br />
<!-- <Option key="id" value="validation"/> --> <br />
</[[Config:SampleEvaluator|SampleEvaluator]]></source></div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=Config:SUMO&diff=5103Config:SUMO2010-03-25T10:36:06Z<p>Icouckuy: AutoConfig for SUMO 6.2</p>
<hr />
<div>'''Generated for SUMO toolbox version 7.0'''.<br />
''We are well aware that documentation is not always complete and possibly even out of date in some cases. We try to document everything as best we can but much is limited by available time and manpower. We are an university research group after all. The most up to date documentation can always be found (if not here) in the default.xml configuration file and, of course, in the source files. If something is unclear please dont hesitate to [[Reporting problems|ask]].''<br />
== SUMO ==<br />
=== Custom Options ===<br />
Available options:<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<br />
<!--Should a movie be created of the model plots when the toolbox has terminated--><br />
<Option key="createMovie" value="yes"/><br />
<!--The minimum amount of samples alotted to *EACH RUN*, dont stop untill we have at least this many samples--><br />
<Option key="minimumTotalSamples" value="0"/><br />
<!--The maximum amount of samples alotted to *EACH RUN*, stop the run and proceed to the next if this number of samples is exceeded (set to Inf to disable)--><br />
<Option key="maximumTotalSamples" value="1000"/><br />
<!--The amount of time (in minutes) alotted to *EACH RUN*, stop the run and proceed to the next if this number is exceeded (set to Inf to disable)--><br />
<Option key="maximumTime" value="Inf"/><br />
<!--The maximum number of adaptive modeling iterations alotted to *EACH RUN*, stop the run and proceed to the next if this number is exceeded (set to Inf to disable).--><br />
<Option key="maxModelingIterations" value="Inf"/><br />
<!--How should the random number generator files be seeded, 3 options: - default: do nothing, the same seed will be used each time matlab is started - random: random initial state - file: load the state from the file specified by the randomStateFile option--><br />
<Option key="seedRandomState" value="default"/><br />
<!--Stop the main loop if a fatal error occurs in the sample evaluator, if set to false the toolbox will switch to adaptive modeling mode (further sampling is switched off).--><br />
<Option key="stopOnError" value="true"/><br />
<!--Minimum amount of samples that are to be evaluated from the initial sample set before the modeling process starts. This can either be an absolute number (e.g., 23) or a percentage (e.g., 95%)--><br />
<Option key="minimumInitialSamples" value="100%"/><br />
<!--Maximum number of pending samples allowed at any time in the toolbox.--><br />
<Option key="maximumSamples" value="10"/><br />
<!--How many % samples should be at least retrieved every iteration (relative to the number of selected samples): 0 %: just takes what is finished and continue modeling 100 %: wait until all selected samples have been evaluated--><br />
<Option key="minimumAdaptiveSamples" value="0"/><br />
<!--Must samples be checked against the constraints before they are submitted for evaluation?--><br />
<Option key="newSamplesMustSatisfyConstraints" value="yes"/><br />
<!--Must the entire dataset be used in adaptive modeling mode or not? true: only the initial design is evaluated, and is used in adaptive modeling mode false: the entire dataset is loaded immediately and used in adaptive modeling mode--><br />
<Option key="adaptiveModelingInitialDesignOnly" value="no"/></source></div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=Config:Optimizer&diff=5102Config:Optimizer2010-03-25T10:32:15Z<p>Icouckuy: AutoConfig for SUMO 6.2</p>
<hr />
<div>'''Generated for SUMO toolbox version 7.0'''.<br />
''We are well aware that documentation is not always complete and possibly even out of date in some cases. We try to document everything as best we can but much is limited by available time and manpower. We are are a university research group after all. The most up to date documentation can always be found (if not here) in the default.xml configuration file and, of course, in the source files. If something is unclear please dont hesitate to [[Reporting problems|ask]].''<br />
== Optimizer ==<br />
<br />
=== directOptimizer ===<br />
The DIviding RECtangles (DIRECT) optimization technique of Donald D. R. Jones<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:Optimizer|Optimizer]] type="[[Optimizer#DirectOptimizer|DirectOptimizer]]"><br />
<Option key="maxevals" value="1000"/><br />
<Option key="maxits" value="300"/><br />
</[[Config:Optimizer|Optimizer]]><br />
</source><br />
=== patternsearch ===<br />
Matlab Pattern search (patternsearch function of Matlab Direct Search toolbox)<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:Optimizer|Optimizer]] type="[[Optimizer#MatlabPatternSearch|MatlabPatternSearch]]"><br />
<Option key="maxIterations" value="500"/><br />
<Option key="maxFunEvals" value="1000"/><br />
</[[Config:Optimizer|Optimizer]]><br />
</source><br />
=== fminconWithDerivatives ===<br />
Matlab fmincon (Active-set) using derivative information (used for kriging models in SUMO-toolbox)<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<[[Config:Optimizer|Optimizer]] type="[[Optimizer#MatlabOptimizer|MatlabOptimizer]]"><br />
<Option key="gradobj" value="on"/><br />
<Option key="derivativecheck" value="off"/><br />
<Option key="diagnostics" value="off"/><br />
<Option key="algorithm" value="active-set"/><br />
</[[Config:Optimizer|Optimizer]]><br />
</source></div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=Config:Plan&diff=5101Config:Plan2010-03-25T10:31:56Z<p>Icouckuy: AutoConfig for SUMO 6.2</p>
<hr />
<div>'''Generated for SUMO toolbox version 7.0'''.<br />
''We are well aware that documentation is not always complete and possibly even out of date in some cases. We try to document everything as best we can but much is limited by available time and manpower. We are an university research group after all. The most up to date documentation can always be found (if not here) in the default.xml configuration file and, of course, in the source files. If something is unclear please dont hesitate to [[Reporting problems|ask]].''<br />
== Plan ==<br />
=== ContextConfig ===<br />
Default components, these should normally not be changed unless you know what you are doing<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<br />
<!--Default components, these should normally not be changed unless you know what you are doing--><br />
<[[Config:ContextConfig|ContextConfig]]>default</[[Config:ContextConfig|ContextConfig]]></source><br />
=== SUMO ===<br />
Default components, these should normally not be changed unless you know what you are doing<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<br />
<!--Default components, these should normally not be changed unless you know what you are doing--><br />
<[[Config:SUMO|SUMO]]>default</[[Config:SUMO|SUMO]]></source><br />
=== LevelPlot ===<br />
Default components, these should normally not be changed unless you know what you are doing<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<br />
<!--Default components, these should normally not be changed unless you know what you are doing--><br />
<[[Config:LevelPlot|LevelPlot]]>default</[[Config:LevelPlot|LevelPlot]]></source><br />
=== Simulator ===<br />
This is the problem we are going to model, it refers to the name of a project directory in the examples/ folder. It is also possible to specify an absolute path or to specify a particular xml file within a project directory<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<br />
<!--This is the problem we are going to model, it refers to the name of a project directory in the examples/ folder. It is also possible to specify an absolute path or to specify a particular xml file within a project directory--><br />
<[[Config:Simulator|Simulator]]>Math/Academic2DTwice</[[Config:Simulator|Simulator]]></source><br />
=== Run ===<br />
Runs can given a custom name by using the name attribute, a repeat attribute is also possible to repeat a run multiple times. Placeholders available for run names include: #adaptivemodelbuilder# #simulator# #sampleselector# #output# #measure#<br />
<source xmlns:saxon="http://icl.com/saxon" lang="xml"><br />
<br />
<!--Runs can given a custom name by using the name attribute, a repeat attribute is also possible to repeat a run multiple times. Placeholders available for run names include: #adaptivemodelbuilder# #simulator# #sampleselector# #output# #measure#--><br />
<[[Config:Run|Run]] name="" repeat="1"><br />
<!-- Enties listed here override those defined on plan level --><br />
<br />
<!-- What experimental design to use for the very first set of samples --><br />
<[[Config:InitialDesign|InitialDesign]]>lhdWithCornerPoints</[[Config:InitialDesign|InitialDesign]]><br />
<br />
<!--<br />
The method to use for selecting new samples. Again 'default' is an id that refers to a<br />
SampleSelector tag defined below. To switch off sampling simply remove this tag. --><br />
<[[Config:SampleSelector|SampleSelector]]>default</[[Config:SampleSelector|SampleSelector]]><br />
<br />
<!--<br />
How is the simulator implemented (ie, where does the data come from): <br />
- Matlab script (matlab)<br />
- scattered dataset (scatteredDataset), <br />
- local executable or script (local)<br />
- etc<br />
<br />
Make sure this entry matches what is declared in the simulator xml file<br />
in the project directory. For example, it makes no sense to put matlab here if you only<br />
have a scattered dataset to work with.<br />
--><br />
<[[Config:SampleEvaluator|SampleEvaluator]]>matlab</[[Config:SampleEvaluator|SampleEvaluator]]><br />
<br />
<!--<br />
The AdaptiveModelBuilder specifies the model type and the hyperparameter optimization<br />
algorithm (= the algorithm to choose the model parameters, also referred to as the<br />
modeling algorithm or model builder) to use. The default value 'kriging' refers to Kriging models.<br />
'kriging' is an id that refers to an AdaptiveModelBuilder tag that is defined below.<br />
--><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]>kriging</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
<br />
<!-- How the quality of a model is assesed is determined by one or more Measures. You can try different combinations<br />
of measures by specifying different measure tags. It is the measure score(s) that drive the model parameter optimization.<br />
We recommend you do not use more than one measure unless you know what you are doing.<br />
<br />
If the use attribute is set to 'off' then the measure score is printed and logged, but is not used in the modeling itself.<br />
More examples of measures are shown below.<br />
--><br />
<br />
<[[Config:Measure|Measure]] type="[[Measure#CrossValidation|CrossValidation]]" target="0.01" errorFcn="rootRelativeSquareError" use="on"/><br />
<br />
<!-- By default all inputs are modeled. If you want to only model a couple of inputs you can specify an Inputs tag as follows: <br />
<br />
<[[Config:Inputs|Inputs]]><br />
<[[Config:Input|Input]] name="x" /><br />
<[[Config:Input|Input]] name="y" /><br />
// Setting a simulator input to a constant (default is 0):<br />
<[[Config:Input|Input]] name="z" value="14.6"/><br />
</[[Config:Inputs|Inputs]]><br />
--><br />
<br />
<!-- <br />
By default the toolbox will model every single output using a separate model. If you want to change this<br />
e.g., you only want to model a specific output, or you want to use different settings for each output; then you<br />
can specify an Outputs tag.<br />
<br />
The following is an example for the Academic2DTwice problem used in this file. Remember that if you change<br />
the problem you are modeling, you will have to change this section too.<br />
--><br />
<[[Config:Outputs|Outputs]]><br />
<[[Config:Output|Output]] name="out"><br />
<!--<br />
You can specify output specific configuration here<br />
<br />
<[[Config:SampleSelector|SampleSelector]]>lola</[[Config:SampleSelector|SampleSelector]]><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]>rational</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
<[[Config:Measure|Measure]] type="[[Measure#CrossValidation|CrossValidation]]" target=".01" errorFcn="meanSquareError" use="on" /><br />
--><br />
</[[Config:Output|Output]]><br />
<br />
<[[Config:Output|Output]] name="outinverse"><br />
<!--<br />
<[[Config:SampleSelector|SampleSelector]]>delaunay</[[Config:SampleSelector|SampleSelector]]><br />
<[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]>rbf</[[Config:AdaptiveModelBuilder|AdaptiveModelBuilder]]><br />
<[[Config:Measure|Measure]] type="[[Measure#ValidationSet|ValidationSet]]" target=".05" use="on" /><br />
--><br />
</[[Config:Output|Output]]><br />
<br />
</[[Config:Outputs|Outputs]]><br />
<br />
<!-- <br />
This is a more complex example of how you can have different configurations per output.<br />
--><br />
<!--<br />
<[[Config:Outputs|Outputs]]><br />
<br />
* Model the modulus of complex output S22 using cross-validation and the default model<br />
builder and sample selector.<br />
<br />
<[[Config:Output|Output]] name="S22" complexHandling="modulus"><br />
<[[Config:Measure|Measure]] type="[[Measure#CrossValidation|CrossValidation]]" target=".05" /><br />
</[[Config:Output|Output]]><br />
<br />
* Model the real part of complex output S22, but introduce some normally-distributed noise<br />
(variance .01 by default).<br />
<br />
<[[Config:Output|Output]] name="S22" complexHandling="real"><br />
<[[Config:Measure|Measure]] type="[[Measure#CrossValidation|CrossValidation]]" target=".05" /><br />
* for other types of modifiers see the datamodifiers subdirectory<br />
<[[Config:Modifier|Modifier]] type="[[Modifier#Noise|Noise]]" /><br />
</[[Config:Output|Output]]><br />
--><br />
<br />
<!-- <br />
More complex examples of how you can use measures:<br />
<br />
* 5-fold crossvalidation (warning expensive on some model types!)<br />
<[[Config:Measure|Measure]] type="[[Measure#CrossValidation|CrossValidation]]" target=".001" use="on"><br />
<Option key="folds" value="5"/><br />
</[[Config:Measure|Measure]]> <br />
<br />
* Using a validation set, the size taken as 20% of the available samples<br />
<[[Config:Measure|Measure]] type="[[Measure#ValidationSet|ValidationSet]]" target=".001" errorFcn="meanAbsoluteError"><br />
<Option key="percentUsed" value="20"/><br />
</[[Config:Measure|Measure]]><br />
<br />
* Using a validation set defined in an external file (scattered data)<br />
<[[Config:Measure|Measure]] type="[[Measure#ValidationSet|ValidationSet]]" target=".001"><br />
* the validation set come from a file<br />
<Option key="type" value="file"/><br />
* the test data is scattered data so we need a scattered sample evaluator<br />
to load the data and evaluate the points. The filename is taken from the<br />
<[[Config:ScatteredDataFile|ScatteredDataFile]]> tag in the simulator xml file.<br />
Optionally you can specify an option with key "id" to specify a specifc<br />
dataset if there is more than one choice.<br />
<[[Config:SampleEvaluator|SampleEvaluator]]<br />
type="ibbt.sumo.sampleevaluators.datasets.ScatteredDatasetSampleEvaluator"/><br />
</[[Config:Measure|Measure]]><br />
<br />
* Used for testing optimization problems<br />
* Calculates the (relative) error between the current minimum and a known minimum.<br />
Often one uses this just as a stopping criterion for benchmarking problems.<br />
* trueValue: a known global minimum<br />
<[[Config:Measure|Measure]] type="[[Measure#TestMinimum|TestMinimum]]" errorFcn="relativeError" trueValue="-5.0" target="0.1" use="on" /> <br />
--><br />
</[[Config:Run|Run]]></source></div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=OoDACE:ooDACE_toolbox&diff=5089OoDACE:ooDACE toolbox2010-02-24T14:29:03Z<p>Icouckuy: </p>
<hr />
<div>== Introduction ==<br />
<br />
The blindDACE toolbox is a versatile Matlab toolbox that implements the popular Gaussian Process based kriging surrogate models. Kriging is in particular popular for approximating (and optimizing) deterministic computer experiments. Given a dataset the toolbox automatically fits a kriging surrogate model to it. Afterwards the kriging surrogate can be fully exploited instead of the (probably more expensive) simulation code.<br />
<br />
The toolbox is aimed for solving complex applications (expensive simulation codes, physical experiments, ...) and for researching new kriging extensions and techniques.<br />
<br />
== Download ==<br />
<br />
See: [http://sumo.intec.ugent.be/?q=blindDACE download page]<br />
<br />
== Quick start guide ==<br />
<br />
'''IMPORTANT: Before the toolbox can be used you have to include the toolbox in Matlab's search path. You can do this manually by running startup or if Matlab is started in the root toolbox directory then startup will be run automatically.'''<br />
<br />
<source lang="matlab"><br />
startup<br />
</source><br />
<br />
Now the toolbox is ready to be used. The blindDACE toolbox is designed in an object oriented (OO) fashion.<br />
It is strongly recommended to exploit the OO design directly, i.e., use the Kriging and Optimizer matlab classes.<br />
However, for convenience wrapper scripts (dacefit, predictor) are provided that simulate the DACE toolbox interface (see [[#DACE toolbox interface|wrapper scripts]] for more information).<br />
<br />
Assuming we want to fit a dataset of n samples in d dimensions.<br />
<b>samples</b> holds the input parameters nXd array (each row is one observation) and <b>values</b> is the corresponding nX1 array containing the output values.<br />
<b>lb</b> and <b>ub</b> are 1Xd arrays defining the lower bounds and upper bounds, respectively, needed to optimize the hyperparameters. In addition, a start values has to be specified (e.g., <b>theta0</b> is also an 1Xd array)<br />
<br />
The example code to fit the dataset is then:<br />
<source lang="matlab"><br />
...<br />
% Generate kriging options structure<br />
opts = getDefaultOptions();<br />
opts.hpBounds = [lb ; ub]; % hyperparameter optimization bounds<br />
<br />
% configure the optimization algorithm (only one optimizer is included)<br />
% the Matlab Optimization toolbox is REQUIRED<br />
optimopts.GradObj = 'on';<br />
optimopts.DerivativeCheck = 'off';<br />
optimopts.Diagnostics = 'off';<br />
optimopts.Algorithm = 'active-set';<br />
opts.hpOptimizer = MatlabOptimizer( dim, 1, optimopts );<br />
<br />
% build and fit Kriging object<br />
k = Kriging( opts, theta0, 'regpoly0', @corrgauss );<br />
k = k.fit( samples, values );<br />
<br />
% k represents the approximation and can now be used, e.g.,<br />
[y mse] = k.predict( [1 2] )<br />
...<br />
</source><br />
<br />
See the included demo.m script for more example code on how to use the blindDACE toolbox (including more advanced features such as using blind kriging or how to use regression instead of interpolation). For more information on the classes and their methods please refer to the source files.<br />
<br />
== DACE toolbox interface ==<br />
<br />
The blindDACE toolbox provides two scripts dacefit.m and predictor.m that simulate the behavior of the DACE toolbox ([http://www2.imm.dtu.dk/~hbn/dace/]). Note, that full compatibility between blindDACE and the DACE toolbox is not provided. The scripts merely aim to ease the transition from the DACE toolbox to blindDACE.<br />
<br />
Example code:<br />
<source lang="matlab"><br />
krige = dacefit(samples, values, 'regpoly0', 'corrgauss', theta0, lb, ub )<br />
y = predictor([1 2], krige)<br />
</source><br />
<br />
Obviously, a lot less code is used to copy the setup described above. However, less code means less flexibility (e.g., blind kriging and regression kriging are not available using the wrapper scripts). Hence, it is suggested to learn the object oriented interface of blindDACE and use that instead.<br />
<br />
== Contribute ==<br />
<br />
Suggestions on how to improve the blindDACE toolbox are always welcome. For more information please see the [[Feedback | feedback]] page.</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=OoDACE:ooDACE_toolbox&diff=5087OoDACE:ooDACE toolbox2010-02-24T11:01:11Z<p>Icouckuy: /* Contribute */</p>
<hr />
<div>== Introduction ==<br />
<br />
The blindDACE toolbox is a versatile Matlab toolbox that implements the popular Gaussian Process based kriging surrogate models. Kriging is in particular popular for approximating (and optimizing) deterministic computer experiments. Given a dataset the toolbox automatically fits a kriging surrogate model to it. Afterwards the kriging surrogate can be fully exploited instead of the (probably more expensive) simulation code.<br />
<br />
The toolbox is aimed for solving complex applications (expensive simulation codes, physical experiments, ...) and for researching new kriging extensions and techniques.<br />
<br />
== Download ==<br />
<br />
See: [http://sumo.intec.ugent.be/?q=blindDACE download page]<br />
<br />
== Quick start guide ==<br />
<br />
'''IMPORTANT: Before the toolbox can be used you have to include the toolbox in Matlab's path. You can do this manually by running startup or if Matlab is started in the root toolbox directory then startup will be run automatically.'''<br />
<br />
<source lang="matlab"><br />
startup<br />
</source><br />
<br />
<br />
Now the toolbox is ready to be used. The blindDACE toolbox is designed in an object oriented (OO) fashion.<br />
It is strongly recommended to exploit the OO design directly, i.e., use the Kriging and Optimizer matlab classes.<br />
However, for convenience wrapper scripts (dacefit, predictor) are provided that simulate the DACE toolbox interface (see [[#DACE toolbox interface|wrapper scripts]] for more information).<br />
<br />
Assuming we want to fit a dataset of n samples in d dimensions.<br />
<b>samples</b> holds the input parameters nXd array (each row is one observation) and <b>values</b> is the corresponding nX1 array containing the output values.<br />
<b>lb</b> and <b>ub</b> are 1Xd arrays defining the lower bounds and upper bounds, respectively, needed to optimize the hyperparameters. In addition, a start values has to be specified (e.g., <b>theta0</b> is also an 1Xd array)<br />
<br />
The example code to fit the dataset follows:<br />
<source lang="matlab"><br />
...<br />
% Generate kriging options structure<br />
opts = getDefaultOptions();<br />
opts.hpBounds = [lb ; ub]; % hyperparameter optimization bounds<br />
<br />
% configure the optimization algorithm (only one optimizer is included)<br />
% the Matlab Optimization toolbox is REQUIRED<br />
optimopts.GradObj = 'on';<br />
optimopts.DerivativeCheck = 'off';<br />
optimopts.Diagnostics = 'off';<br />
optimopts.Algorithm = 'active-set';<br />
opts.hpOptimizer = MatlabOptimizer( dim, 1, optimopts );<br />
<br />
% build and fit Kriging object<br />
k = Kriging( opts, theta0, 'regpoly0', @corrgauss );<br />
k = k.fit( samples, values );<br />
<br />
% k represents the approximation and can now be used, e.g.,<br />
[y mse] = k.predict( [1 2] )<br />
...<br />
</source><br />
<br />
See the included demo.m script for more example code on how to use the blindDACE toolbox (including more advances features such as using blind kriging or how to use regression instead of interpolation). For more information on the classes and their methods please refer to the source files.<br />
<br />
== DACE toolbox interface ==<br />
<br />
The blindDACE toolbox provides two scripts dacefit.m and predictor.m that simulate the behavior of the DACE toolbox ([http://www2.imm.dtu.dk/~hbn/dace/]). Note, that full compatibility between blindDACE and the DACE toolbox is not provided. The scripts merely aim to ease the transition from the DACE toolbox to blindDACE.<br />
<br />
Example code:<br />
<source lang="matlab"><br />
krige = dacefit(samples, values, 'regpoly0', 'corrgauss', theta0, lb, ub )<br />
y = predictor([1 2], krige)<br />
</source><br />
<br />
Obviously, a lot less code is used to copy the setup described above. However, less code means less flexibility (e.g., blind kriging and regression kriging is not available using the wrapper scripts). Hence, it is suggested to learn the object oriented interface of blindDACE and use that instead.<br />
<br />
== Contribute ==<br />
<br />
Suggestions on how to improve the blindDACE toolbox are always welcome. For more information please see the [[Feedback | feedback]] page.</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=OoDACE:ooDACE_toolbox&diff=5086OoDACE:ooDACE toolbox2010-02-24T11:00:58Z<p>Icouckuy: /* Contribute */</p>
<hr />
<div>== Introduction ==<br />
<br />
The blindDACE toolbox is a versatile Matlab toolbox that implements the popular Gaussian Process based kriging surrogate models. Kriging is in particular popular for approximating (and optimizing) deterministic computer experiments. Given a dataset the toolbox automatically fits a kriging surrogate model to it. Afterwards the kriging surrogate can be fully exploited instead of the (probably more expensive) simulation code.<br />
<br />
The toolbox is aimed for solving complex applications (expensive simulation codes, physical experiments, ...) and for researching new kriging extensions and techniques.<br />
<br />
== Download ==<br />
<br />
See: [http://sumo.intec.ugent.be/?q=blindDACE download page]<br />
<br />
== Quick start guide ==<br />
<br />
'''IMPORTANT: Before the toolbox can be used you have to include the toolbox in Matlab's path. You can do this manually by running startup or if Matlab is started in the root toolbox directory then startup will be run automatically.'''<br />
<br />
<source lang="matlab"><br />
startup<br />
</source><br />
<br />
<br />
Now the toolbox is ready to be used. The blindDACE toolbox is designed in an object oriented (OO) fashion.<br />
It is strongly recommended to exploit the OO design directly, i.e., use the Kriging and Optimizer matlab classes.<br />
However, for convenience wrapper scripts (dacefit, predictor) are provided that simulate the DACE toolbox interface (see [[#DACE toolbox interface|wrapper scripts]] for more information).<br />
<br />
Assuming we want to fit a dataset of n samples in d dimensions.<br />
<b>samples</b> holds the input parameters nXd array (each row is one observation) and <b>values</b> is the corresponding nX1 array containing the output values.<br />
<b>lb</b> and <b>ub</b> are 1Xd arrays defining the lower bounds and upper bounds, respectively, needed to optimize the hyperparameters. In addition, a start values has to be specified (e.g., <b>theta0</b> is also an 1Xd array)<br />
<br />
The example code to fit the dataset follows:<br />
<source lang="matlab"><br />
...<br />
% Generate kriging options structure<br />
opts = getDefaultOptions();<br />
opts.hpBounds = [lb ; ub]; % hyperparameter optimization bounds<br />
<br />
% configure the optimization algorithm (only one optimizer is included)<br />
% the Matlab Optimization toolbox is REQUIRED<br />
optimopts.GradObj = 'on';<br />
optimopts.DerivativeCheck = 'off';<br />
optimopts.Diagnostics = 'off';<br />
optimopts.Algorithm = 'active-set';<br />
opts.hpOptimizer = MatlabOptimizer( dim, 1, optimopts );<br />
<br />
% build and fit Kriging object<br />
k = Kriging( opts, theta0, 'regpoly0', @corrgauss );<br />
k = k.fit( samples, values );<br />
<br />
% k represents the approximation and can now be used, e.g.,<br />
[y mse] = k.predict( [1 2] )<br />
...<br />
</source><br />
<br />
See the included demo.m script for more example code on how to use the blindDACE toolbox (including more advances features such as using blind kriging or how to use regression instead of interpolation). For more information on the classes and their methods please refer to the source files.<br />
<br />
== DACE toolbox interface ==<br />
<br />
The blindDACE toolbox provides two scripts dacefit.m and predictor.m that simulate the behavior of the DACE toolbox ([http://www2.imm.dtu.dk/~hbn/dace/]). Note, that full compatibility between blindDACE and the DACE toolbox is not provided. The scripts merely aim to ease the transition from the DACE toolbox to blindDACE.<br />
<br />
Example code:<br />
<source lang="matlab"><br />
krige = dacefit(samples, values, 'regpoly0', 'corrgauss', theta0, lb, ub )<br />
y = predictor([1 2], krige)<br />
</source><br />
<br />
Obviously, a lot less code is used to copy the setup described above. However, less code means less flexibility (e.g., blind kriging and regression kriging is not available using the wrapper scripts). Hence, it is suggested to learn the object oriented interface of blindDACE and use that instead.<br />
<br />
== Contribute ==<br />
<br />
Suggestions on how to improve the blindDACE toolbox are always welcome. For more information please see the [[Feedback feedback]] page.</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=OoDACE:ooDACE_toolbox&diff=5085OoDACE:ooDACE toolbox2010-02-24T10:59:58Z<p>Icouckuy: </p>
<hr />
<div>== Introduction ==<br />
<br />
The blindDACE toolbox is a versatile Matlab toolbox that implements the popular Gaussian Process based kriging surrogate models. Kriging is in particular popular for approximating (and optimizing) deterministic computer experiments. Given a dataset the toolbox automatically fits a kriging surrogate model to it. Afterwards the kriging surrogate can be fully exploited instead of the (probably more expensive) simulation code.<br />
<br />
The toolbox is aimed for solving complex applications (expensive simulation codes, physical experiments, ...) and for researching new kriging extensions and techniques.<br />
<br />
== Download ==<br />
<br />
See: [http://sumo.intec.ugent.be/?q=blindDACE download page]<br />
<br />
== Quick start guide ==<br />
<br />
'''IMPORTANT: Before the toolbox can be used you have to include the toolbox in Matlab's path. You can do this manually by running startup or if Matlab is started in the root toolbox directory then startup will be run automatically.'''<br />
<br />
<source lang="matlab"><br />
startup<br />
</source><br />
<br />
<br />
Now the toolbox is ready to be used. The blindDACE toolbox is designed in an object oriented (OO) fashion.<br />
It is strongly recommended to exploit the OO design directly, i.e., use the Kriging and Optimizer matlab classes.<br />
However, for convenience wrapper scripts (dacefit, predictor) are provided that simulate the DACE toolbox interface (see [[#DACE toolbox interface|wrapper scripts]] for more information).<br />
<br />
Assuming we want to fit a dataset of n samples in d dimensions.<br />
<b>samples</b> holds the input parameters nXd array (each row is one observation) and <b>values</b> is the corresponding nX1 array containing the output values.<br />
<b>lb</b> and <b>ub</b> are 1Xd arrays defining the lower bounds and upper bounds, respectively, needed to optimize the hyperparameters. In addition, a start values has to be specified (e.g., <b>theta0</b> is also an 1Xd array)<br />
<br />
The example code to fit the dataset follows:<br />
<source lang="matlab"><br />
...<br />
% Generate kriging options structure<br />
opts = getDefaultOptions();<br />
opts.hpBounds = [lb ; ub]; % hyperparameter optimization bounds<br />
<br />
% configure the optimization algorithm (only one optimizer is included)<br />
% the Matlab Optimization toolbox is REQUIRED<br />
optimopts.GradObj = 'on';<br />
optimopts.DerivativeCheck = 'off';<br />
optimopts.Diagnostics = 'off';<br />
optimopts.Algorithm = 'active-set';<br />
opts.hpOptimizer = MatlabOptimizer( dim, 1, optimopts );<br />
<br />
% build and fit Kriging object<br />
k = Kriging( opts, theta0, 'regpoly0', @corrgauss );<br />
k = k.fit( samples, values );<br />
<br />
% k represents the approximation and can now be used, e.g.,<br />
[y mse] = k.predict( [1 2] )<br />
...<br />
</source><br />
<br />
See the included demo.m script for more example code on how to use the blindDACE toolbox (including more advances features such as using blind kriging or how to use regression instead of interpolation). For more information on the classes and their methods please refer to the source files.<br />
<br />
== DACE toolbox interface ==<br />
<br />
The blindDACE toolbox provides two scripts dacefit.m and predictor.m that simulate the behavior of the DACE toolbox ([http://www2.imm.dtu.dk/~hbn/dace/]). Note, that full compatibility between blindDACE and the DACE toolbox is not provided. The scripts merely aim to ease the transition from the DACE toolbox to blindDACE.<br />
<br />
Example code:<br />
<source lang="matlab"><br />
krige = dacefit(samples, values, 'regpoly0', 'corrgauss', theta0, lb, ub )<br />
y = predictor([1 2], krige)<br />
</source><br />
<br />
Obviously, a lot less code is used to copy the setup described above. However, less code means less flexibility (e.g., blind kriging and regression kriging is not available using the wrapper scripts). Hence, it is suggested to learn the object oriented interface of blindDACE and use that instead.<br />
<br />
== Contribute ==<br />
<br />
Suggestions on how to improve the blindDACE toolbox are always welcome. For a detailed list of feedback points, please see the [[http://sumowiki.intec.ugent.be/index.php/Feedback feedback]] page.</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=OoDACE:ooDACE_toolbox&diff=5084OoDACE:ooDACE toolbox2010-02-24T10:36:13Z<p>Icouckuy: </p>
<hr />
<div>== Introduction ==<br />
<br />
The blindDACE toolbox is a versatile Matlab toolbox that implements the popular Gaussian Process based kriging surrogate models. Kriging is in particular popular for approximating (and optimizing) deterministic computer experiments. Given a dataset the toolbox automatically fits a kriging surrogate model to it. Afterwards the kriging surrogate can be fully exploited instead of the (probably more expensive) simulation code.<br />
<br />
The toolbox is aimed for solving complex applications (expensive simulation codes, physical experiments, ...) and for researching new kriging extensions and techniques.<br />
<br />
== Download ==<br />
<br />
See: [http://sumo.intec.ugent.be/?q=blindDACE download page]<br />
<br />
== Quick start guide ==<br />
<br />
'''IMPORTANT: Before the toolbox can be used you have to include the toolbox in Matlab's path. You can do this manually by running startup or if Matlab is started in the root toolbox directory then startup will be run automatically.'''<br />
<br />
<source lang="matlab"><br />
startup<br />
</source><br />
<br />
<br />
Now the toolbox is ready to be used. The blindDACE toolbox is designed in an object oriented (OO) fashion.<br />
It is strongly recommended to exploit the OO design directly, i.e., use the Kriging and Optimizer matlab classes.<br />
However, for convenience wrapper scripts (dacefit, predictor) are provided that simulate the DACE toolbox interface (see [[#DACE toolbox interface|wrapper scripts]] for more information).<br />
<br />
Assuming we want to fit a dataset of n samples in d dimensions.<br />
<b>samples</b> holds the input parameters nXd array (each row is one observation) and <b>values</b> is the corresponding nX1 array containing the output values.<br />
<b>lb</b> and <b>ub</b> are 1Xd arrays defining the lower bounds and upper bounds, respectively, needed to optimize the hyperparameters. In addition, a start values has to be specified (e.g., <b>theta0</b> is also an 1Xd array)<br />
<br />
The example code to fit the dataset follows:<br />
<source lang="matlab"><br />
...<br />
% Generate kriging options structure<br />
opts = getDefaultOptions();<br />
opts.hpBounds = [lb ; ub]; % hyperparameter optimization bounds<br />
<br />
% configure the optimization algorithm (only one optimizer is included)<br />
% the Matlab Optimization toolbox is REQUIRED<br />
optimopts.GradObj = 'on';<br />
optimopts.DerivativeCheck = 'off';<br />
optimopts.Diagnostics = 'off';<br />
optimopts.Algorithm = 'active-set';<br />
opts.hpOptimizer = MatlabOptimizer( dim, 1, optimopts );<br />
<br />
% build and fit Kriging object<br />
k = Kriging( opts, theta0, 'regpoly0', @corrgauss );<br />
k = k.fit( samples, values );<br />
<br />
% k represents the approximation and can now be used, e.g.,<br />
[y mse] = k.predict( [1 2] )<br />
...<br />
</source><br />
<br />
See the included demo.m script for more example code on how to use the blindDACE toolbox (including more advances features such as using blind kriging or how to use regression instead of interpolation). For more information on the classes and their methods please refer to the source files.<br />
<br />
== DACE toolbox interface ==<br />
<br />
The blindDACE toolbox provides two scripts dacefit.m and predictor.m that simulate the behavior of the DACE toolbox ([http://www2.imm.dtu.dk/~hbn/dace/]). Note, that full compatibility between blindDACE and the DACE toolbox is not provided. The scripts merely aim to ease the transition from the DACE toolbox to blindDACE.<br />
<br />
Example code:<br />
<source lang="matlab"><br />
krige = dacefit(samples, values, 'regpoly0', 'corrgauss', theta0, lb, ub )<br />
y = predictor([1 2], krige)<br />
</source><br />
<br />
Obviously, a lot less code is used to copy the setup described above. However, less code means less flexibility (e.g., blind kriging and regression kriging is not available using the wrapper scripts). Hence, it is suggested to learn the object oriented interface of blindDACE and use that instead.<br />
<br />
== Contribute ==<br />
<br />
These bindings are very basic but they work for me. Please improve, extend, provide binaries, and of course contribute back.</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=BlindDACE_toolbox&diff=5083BlindDACE toolbox2010-02-17T14:39:50Z<p>Icouckuy: BlindDACE toolbox moved to BlindDACE:BlindDACE toolbox: own namespace</p>
<hr />
<div>#REDIRECT [[BlindDACE:BlindDACE toolbox]]</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=OoDACE:ooDACE_toolbox&diff=5082OoDACE:ooDACE toolbox2010-02-17T14:39:50Z<p>Icouckuy: BlindDACE toolbox moved to BlindDACE:BlindDACE toolbox: own namespace</p>
<hr />
<div>== Introduction ==<br />
<br />
he blindDACE toolbox is a versatile Matlab toolbox that implements the popular Gaussian Process based kriging surrogate models. Kriging is in particular popular for approximating (and optimizing) deterministic computer experiments. Given a dataset the toolbox automatically fits a kriging surrogate model to it. Afterwards the kriging surrogate can be fully exploited instead of the (probably more expensive) simulation code.<br />
<br />
The toolbox is aimed for solving complex applications (expensive simulation codes, physical experiments, ...) and for researching new kriging extensions and techniques.<br />
<br />
== Download ==<br />
<br />
See: [http://sumo.intec.ugent.be/?q=blindDACE download page]<br />
<br />
== Quick start guide ==<br />
<br />
'''IMPORTANT: Before the toolbox can be used you have to include the toolbox in Matlab's path. You can do this manually by running startup or if Matlab is started in the root toolbox directory then startup will be run automatically.'''<br />
<br />
<source lang="matlab"><br />
startup<br />
</source><br />
<br />
<br />
Now the toolbox is ready to be used. The blindDACE toolbox is designed in an object oriented (OO) fashion.<br />
It is strongly recommended to exploit the OO design directly, i.e., use the Kriging and Optimizer matlab classes.<br />
However, for convenience wrapper scripts (dacefit, predictor) are provided that simulate the DACE toolbox interface (see [[#DACE toolbox interface|wrapper scripts]] for more information).<br />
<br />
Assuming we want to fit a dataset of n samples in d dimensions.<br />
<b>samples</b> holds the input parameters nXd array (each row is one observation) and <b>values</b> is the corresponding nX1 array containing the output values.<br />
<b>lb</b> and <b>ub</b> are 1Xd arrays defining the lower bounds and upper bounds, respectively, needed to optimize the hyperparameters. In addition, a start values has to be specified (e.g., <b>theta0</b> is also an 1Xd array)<br />
<br />
The example code to fit the dataset follows:<br />
<source lang="matlab"><br />
...<br />
% Generate kriging options structure<br />
opts = getDefaultOptions();<br />
opts.hpBounds = [lb ; ub]; % hyperparameter optimization bounds<br />
<br />
% configure the optimization algorithm (only one optimizer is included)<br />
% the Matlab Optimization toolbox is REQUIRED<br />
optimopts.GradObj = 'on';<br />
optimopts.DerivativeCheck = 'off';<br />
optimopts.Diagnostics = 'off';<br />
optimopts.Algorithm = 'active-set';<br />
opts.hpOptimizer = MatlabOptimizer( dim, 1, optimopts );<br />
<br />
% build and fit Kriging object<br />
k = Kriging( opts, theta0, 'regpoly0', @corrgauss );<br />
k = k.fit( samples, values );<br />
<br />
% k represents the approximation and can now be used, e.g.,<br />
[y mse] = k.predict( [1 2] )<br />
...<br />
</source><br />
<br />
See the included demo.m script for more example code on how to use the blindDACE toolbox (including more advances features such as using blind kriging or how to use regression instead of interpolation). For more information on the classes and their methods please refer to the source files.<br />
<br />
== DACE toolbox interface ==<br />
<br />
The blindDACE toolbox provides two scripts dacefit.m and predictor.m that simulate the behavior of the DACE toolbox ([http://www2.imm.dtu.dk/~hbn/dace/]). Note, that full compatibility between blindDACE and the DACE toolbox is not provided. The scripts merely aim to ease the transition from the DACE toolbox to blindDACE.<br />
<br />
Example code:<br />
<source lang="matlab"><br />
krige = dacefit(samples, values, 'regpoly0', 'corrgauss', theta0, lb, ub )<br />
y = predictor([1 2], krige)<br />
</source><br />
<br />
Obviously, a lot less code is used to copy the setup described above. However, less code means less flexibility (e.g., blind kriging and regression kriging is not available using the wrapper scripts). Hence, it is suggested to learn the object oriented interface of blindDACE and use that instead.<br />
<br />
== Contribute ==<br />
<br />
These bindings are very basic but they work for me. Please improve, extend, provide binaries, and of course contribute back.</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=OoDACE:_License_terms&diff=5081OoDACE: License terms2010-02-17T14:38:59Z<p>Icouckuy: </p>
<hr />
<div>[[Image:osilogo.jpg|90px|right|Open Source Initiative]]<br />
<br />
The blindDACE Toolbox is available under a dual license model. <br />
For '''non-commercial''' use, the toolbox is available under the [http://www.fsf.org/licensing/licenses/agpl-3.0.html GNU Affero General Public License version 3] (AGPLv3), an [http://www.opensource.org/ OSI] approved [http://en.wikipedia.org/wiki/Open_source open source] license. <br />
For use in a '''commercial''' setting, a commercial license must be obtained.<br />
<br />
In addition we require that any reference to the blindDACE Toolbox be accompanied by the [[Citing|corresponding publication]].<br />
<br />
== License terms ==<br />
<br />
For '''non-commercial''' use, this program is free software; you can redistribute it and/or modify it under<br />
the terms of the GNU Affero General Public License version 3 as published by the Free Software Foundation.<br />
<br />
This program is distributed in the hope that it will be useful, but WITHOUT ANY<br />
WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A<br />
PARTICULAR PURPOSE. See the GNU Affero General Public License for more details.<br />
<br />
You should have received a copy of the GNU Affero General Public License along<br />
with this program; if not, see http://www.gnu.org/licenses or write to the Free<br />
Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA<br />
02110-1301 USA, or download the license from the following URL:<br />
<br />
[http://www.fsf.org/licensing/licenses/agpl-3.0.html http://www.fsf.org/licensing/licenses/agpl-3.0.html]<br />
<br />
In accordance with Section 7(b) of the GNU Affero General Public License, these<br />
Appropriate Legal Notices must retain the display of the "blindDACE Toolbox" text and<br />
homepage. In addition, when mentioning the program in written work, reference<br />
must be made to the [[Citing|corresponding publication]].<br />
<br />
You can be released from these requirements by purchasing a commercial license.<br />
Buying such a license is in most cases mandatory as soon as you develop<br />
commercial activities involving the blindDACE Toolbox software. Commercial activities<br />
include: consultancy services or using the blindDACE Toolbox in commercial projects <br />
(standalone, on a server, through a webservice or other remote access technology).<br />
<br />
For details about a '''commercial''' license please [[Contact]] us.</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=OoDACE:_License_terms&diff=5080OoDACE: License terms2010-02-17T14:38:23Z<p>Icouckuy: New page: Open Source Initiative The blindDACE Toolbox is available under a dual license model. For '''non-commercial''' use, the toolbox is available under the [h...</p>
<hr />
<div>[[Image:osilogo.jpg|90px|right|Open Source Initiative]]<br />
<br />
The blindDACE Toolbox is available under a dual license model. <br />
For '''non-commercial''' use, the toolbox is available under the [http://www.fsf.org/licensing/licenses/agpl-3.0.html GNU Affero General Public License version 3] (AGPLv3), an [http://www.opensource.org/ OSI] approved [http://en.wikipedia.org/wiki/Open_source open source] license. <br />
For use in a '''commercial''' setting, a commercial license must be obtained.<br />
<br />
In addition we require that any reference to the blindDACE Toolbox be accompanied by the [[Citing|corresponding publication]].<br />
<br />
== License terms ==<br />
<br />
For '''non-commercial''' use, this program is free software; you can redistribute it and/or modify it under<br />
the terms of the GNU Affero General Public License version 3 as published by the Free Software Foundation.<br />
<br />
This program is distributed in the hope that it will be useful, but WITHOUT ANY<br />
WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A<br />
PARTICULAR PURPOSE. See the GNU Affero General Public License for more details.<br />
<br />
You should have received a copy of the GNU Affero General Public License along<br />
with this program; if not, see http://www.gnu.org/licenses or write to the Free<br />
Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA<br />
02110-1301 USA, or download the license from the following URL:<br />
<br />
[http://www.fsf.org/licensing/licenses/agpl-3.0.html http://www.fsf.org/licensing/licenses/agpl-3.0.html]<br />
<br />
In accordance with Section 7(b) of the GNU Affero General Public License, these<br />
Appropriate Legal Notices must retain the display of the "blindDACE Toolbox" text and<br />
homepage. In addition, when mentioning the program in written work, reference<br />
must be made to the [[Citing|corresponding publication]].<br />
<br />
You can be released from these requirements by purchasing a commercial license.<br />
Buying such a license is in most cases mandatory as soon as you develop<br />
commercial activities involving the SUMO Toolbox software. Commercial activities<br />
include: consultancy services or using the SUMO Toolbox in commercial projects <br />
(standalone, on a server, through a webservice or other remote access technology).<br />
<br />
For details about a '''commercial''' license please [[Contact]] us.</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=OoDACE:ooDACE_toolbox&diff=5076OoDACE:ooDACE toolbox2010-02-09T10:22:43Z<p>Icouckuy: /* Quick start guide */</p>
<hr />
<div>== Introduction ==<br />
<br />
he blindDACE toolbox is a versatile Matlab toolbox that implements the popular Gaussian Process based kriging surrogate models. Kriging is in particular popular for approximating (and optimizing) deterministic computer experiments. Given a dataset the toolbox automatically fits a kriging surrogate model to it. Afterwards the kriging surrogate can be fully exploited instead of the (probably more expensive) simulation code.<br />
<br />
The toolbox is aimed for solving complex applications (expensive simulation codes, physical experiments, ...) and for researching new kriging extensions and techniques.<br />
<br />
== Download ==<br />
<br />
See: [http://sumo.intec.ugent.be/?q=blindDACE download page]<br />
<br />
== Quick start guide ==<br />
<br />
'''IMPORTANT: Before the toolbox can be used you have to include the toolbox in Matlab's path. You can do this manually by running startup or if Matlab is started in the root toolbox directory then startup will be run automatically.'''<br />
<br />
<source lang="matlab"><br />
startup<br />
</source><br />
<br />
<br />
Now the toolbox is ready to be used. The blindDACE toolbox is designed in an object oriented (OO) fashion.<br />
It is strongly recommended to exploit the OO design directly, i.e., use the Kriging and Optimizer matlab classes.<br />
However, for convenience wrapper scripts (dacefit, predictor) are provided that simulate the DACE toolbox interface (see [[#DACE toolbox interface|wrapper scripts]] for more information).<br />
<br />
Assuming we want to fit a dataset of n samples in d dimensions.<br />
<b>samples</b> holds the input parameters nXd array (each row is one observation) and <b>values</b> is the corresponding nX1 array containing the output values.<br />
<b>lb</b> and <b>ub</b> are 1Xd arrays defining the lower bounds and upper bounds, respectively, needed to optimize the hyperparameters. In addition, a start values has to be specified (e.g., <b>theta0</b> is also an 1Xd array)<br />
<br />
The example code to fit the dataset follows:<br />
<source lang="matlab"><br />
...<br />
% Generate kriging options structure<br />
opts = getDefaultOptions();<br />
opts.hpBounds = [lb ; ub]; % hyperparameter optimization bounds<br />
<br />
% configure the optimization algorithm (only one optimizer is included)<br />
% the Matlab Optimization toolbox is REQUIRED<br />
optimopts.GradObj = 'on';<br />
optimopts.DerivativeCheck = 'off';<br />
optimopts.Diagnostics = 'off';<br />
optimopts.Algorithm = 'active-set';<br />
opts.hpOptimizer = MatlabOptimizer( dim, 1, optimopts );<br />
<br />
% build and fit Kriging object<br />
k = Kriging( opts, theta0, 'regpoly0', @corrgauss );<br />
k = k.fit( samples, values );<br />
<br />
% k represents the approximation and can now be used, e.g.,<br />
[y mse] = k.predict( [1 2] )<br />
...<br />
</source><br />
<br />
See the included demo.m script for more example code on how to use the blindDACE toolbox (including more advances features such as using blind kriging or how to use regression instead of interpolation). For more information on the classes and their methods please refer to the source files.<br />
<br />
== DACE toolbox interface ==<br />
<br />
The blindDACE toolbox provides two scripts dacefit.m and predictor.m that simulate the behavior of the DACE toolbox ([http://www2.imm.dtu.dk/~hbn/dace/]). Note, that full compatibility between blindDACE and the DACE toolbox is not provided. The scripts merely aim to ease the transition from the DACE toolbox to blindDACE.<br />
<br />
Example code:<br />
<source lang="matlab"><br />
krige = dacefit(samples, values, 'regpoly0', 'corrgauss', theta0, lb, ub )<br />
y = predictor([1 2], krige)<br />
</source><br />
<br />
Obviously, a lot less code is used to copy the setup described above. However, less code means less flexibility (e.g., blind kriging and regression kriging is not available using the wrapper scripts). Hence, it is suggested to learn the object oriented interface of blindDACE and use that instead.<br />
<br />
== Contribute ==<br />
<br />
These bindings are very basic but they work for me. Please improve, extend, provide binaries, and of course contribute back.</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=OoDACE:ooDACE_toolbox&diff=5075OoDACE:ooDACE toolbox2010-02-09T10:21:32Z<p>Icouckuy: /* Quick start guide */</p>
<hr />
<div>== Introduction ==<br />
<br />
he blindDACE toolbox is a versatile Matlab toolbox that implements the popular Gaussian Process based kriging surrogate models. Kriging is in particular popular for approximating (and optimizing) deterministic computer experiments. Given a dataset the toolbox automatically fits a kriging surrogate model to it. Afterwards the kriging surrogate can be fully exploited instead of the (probably more expensive) simulation code.<br />
<br />
The toolbox is aimed for solving complex applications (expensive simulation codes, physical experiments, ...) and for researching new kriging extensions and techniques.<br />
<br />
== Download ==<br />
<br />
See: [http://sumo.intec.ugent.be/?q=blindDACE download page]<br />
<br />
== Quick start guide ==<br />
<br />
'''IMPORTANT: Before the toolbox can be used you have to include the toolbox in Matlab's path. You can do this manually by running startup or if Matlab is started in the root toolbox directory then startup will be run automatically.'''<br />
<br />
<source lang="matlab"><br />
startup<br />
</source><br />
<br />
<br />
Now the toolbox is ready to be used. The blindDACE toolbox is designed in an object oriented (OO) fashion.<br />
It is strongly recommended to exploit the OO design directly, i.e., use the Kriging and Optimizer matlab classes.<br />
However, for convenience wrapper scripts (dacefit, predictor) are provided that simulate the DACE toolbox interface (see [[#DACE toolbox interface|wrapper scripts]] for more information).<br />
<br />
Assuming we want to fit a dataset of n samples in d dimensions.<br />
<b>samples</b> holds the input parameters nXd array (each row is one observation) and <b>values</b> is the corresponding nX1 array containing the output values.<br />
<b>lb</b> and <b>ub</b> are 1Xd arrays defining the lower bounds and upper bounds, respectively, needed to optimize the hyperparameters. In addition, a start values has to be specified (e.g., <b>theta0</b> is also an 1Xd array)<br />
<br />
The example code to fit the dataset follows:<br />
<source lang="matlab"><br />
...<br />
% Generate kriging options structure<br />
opts = getDefaultOptions();<br />
opts.hpBounds = [lb ; ub]; % hyperparameter optimization bounds<br />
<br />
% configure the optimization algorithm (only one optimizer is included)<br />
% the Matlab Optimization toolbox is REQUIRED<br />
optimopts.GradObj = 'on';<br />
optimopts.DerivativeCheck = 'off';<br />
optimopts.Diagnostics = 'off';<br />
optimopts.Algorithm = 'active-set';<br />
opts.hpOptimizer = MatlabOptimizer( dim, 1, optimopts );<br />
<br />
% build and fit Kriging object<br />
k = Kriging( opts, theta0, 'regpoly0', @corrgauss );<br />
k = k.fit( samples, values );<br />
<br />
% k represents the approximation and can now be used, e.g.,<br />
[y mse] = k.predict( [1 2] )<br />
...<br />
</source><br />
<br />
See the included demo.m script for more example code on how to use the blindDACE toolbox (including more advances features such as using blind kriging or how to use regression instead of interpolation).<br />
<br />
== DACE toolbox interface ==<br />
<br />
The blindDACE toolbox provides two scripts dacefit.m and predictor.m that simulate the behavior of the DACE toolbox ([http://www2.imm.dtu.dk/~hbn/dace/]). Note, that full compatibility between blindDACE and the DACE toolbox is not provided. The scripts merely aim to ease the transition from the DACE toolbox to blindDACE.<br />
<br />
Example code:<br />
<source lang="matlab"><br />
krige = dacefit(samples, values, 'regpoly0', 'corrgauss', theta0, lb, ub )<br />
y = predictor([1 2], krige)<br />
</source><br />
<br />
Obviously, a lot less code is used to copy the setup described above. However, less code means less flexibility (e.g., blind kriging and regression kriging is not available using the wrapper scripts). Hence, it is suggested to learn the object oriented interface of blindDACE and use that instead.<br />
<br />
== Contribute ==<br />
<br />
These bindings are very basic but they work for me. Please improve, extend, provide binaries, and of course contribute back.</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=OoDACE:ooDACE_toolbox&diff=5074OoDACE:ooDACE toolbox2010-02-09T10:20:32Z<p>Icouckuy: /* Quick start guide */</p>
<hr />
<div>== Introduction ==<br />
<br />
he blindDACE toolbox is a versatile Matlab toolbox that implements the popular Gaussian Process based kriging surrogate models. Kriging is in particular popular for approximating (and optimizing) deterministic computer experiments. Given a dataset the toolbox automatically fits a kriging surrogate model to it. Afterwards the kriging surrogate can be fully exploited instead of the (probably more expensive) simulation code.<br />
<br />
The toolbox is aimed for solving complex applications (expensive simulation codes, physical experiments, ...) and for researching new kriging extensions and techniques.<br />
<br />
== Download ==<br />
<br />
See: [http://sumo.intec.ugent.be/?q=blindDACE download page]<br />
<br />
== Quick start guide ==<br />
<br />
'''IMPORTANT: Before the toolbox can be used you have to include the toolbox in Matlab's path. You can do this manually by running startup or if Matlab is started in the root toolbox directory then startup will be run automatically.'''<br />
<br />
<source lang="matlab"><br />
startup<br />
</source><br />
<br />
<br />
Now the toolbox is ready to be used. The blindDACE toolbox is designed in an object oriented (OO) fashion.<br />
It is strongly recommended to exploit the OO design directly, i.e., use the Kriging and Optimizer matlab classes.<br />
However, for convenience wrapper scripts (dacefit, predictor) are provided that simulate the DACE toolbox interface (see [[#DACE toolbox interface|wrapper scripts]] for more information).<br />
<br />
Assuming we want to fit a dataset of n samples in d dimensions.<br />
<b>samples</b> holds the input parameters nXd array (each row is one observation) and <math>values</math> is the corresponding nX1 array containing the output values.<br />
<math>lb</math> and <math>ub</math> are 1Xd arrays defining the lower bounds and upper bounds, respectively, needed to optimize the hyperparameters. In addition, a start values has to be specified (e.g., <b>theta0</b> is also an 1Xd array)<br />
<br />
The example code to fit the dataset follows:<br />
<source lang="matlab"><br />
...<br />
% Generate kriging options structure<br />
opts = getDefaultOptions();<br />
opts.hpBounds = [lb ; ub]; % hyperparameter optimization bounds<br />
<br />
% configure the optimization algorithm (only one optimizer is included)<br />
% the Matlab Optimization toolbox is REQUIRED<br />
optimopts.GradObj = 'on';<br />
optimopts.DerivativeCheck = 'off';<br />
optimopts.Diagnostics = 'off';<br />
optimopts.Algorithm = 'active-set';<br />
opts.hpOptimizer = MatlabOptimizer( dim, 1, optimopts );<br />
<br />
% build and fit Kriging object<br />
k = Kriging( opts, theta0, 'regpoly0', @corrgauss );<br />
k = k.fit( samples, values );<br />
<br />
% k represents the approximation and can now be used, e.g.,<br />
[y mse] = k.predict( [1 2] )<br />
...<br />
</source><br />
<br />
See the included demo.m script for more example code on how to use the blindDACE toolbox (including more advances features such as using blind kriging or how to use regression instead of interpolation).<br />
<br />
== DACE toolbox interface ==<br />
<br />
The blindDACE toolbox provides two scripts dacefit.m and predictor.m that simulate the behavior of the DACE toolbox ([http://www2.imm.dtu.dk/~hbn/dace/]). Note, that full compatibility between blindDACE and the DACE toolbox is not provided. The scripts merely aim to ease the transition from the DACE toolbox to blindDACE.<br />
<br />
Example code:<br />
<source lang="matlab"><br />
krige = dacefit(samples, values, 'regpoly0', 'corrgauss', theta0, lb, ub )<br />
y = predictor([1 2], krige)<br />
</source><br />
<br />
Obviously, a lot less code is used to copy the setup described above. However, less code means less flexibility (e.g., blind kriging and regression kriging is not available using the wrapper scripts). Hence, it is suggested to learn the object oriented interface of blindDACE and use that instead.<br />
<br />
== Contribute ==<br />
<br />
These bindings are very basic but they work for me. Please improve, extend, provide binaries, and of course contribute back.</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=OoDACE:ooDACE_toolbox&diff=5073OoDACE:ooDACE toolbox2010-02-09T10:13:31Z<p>Icouckuy: /* blindDACE: A versatile kriging Matlab Toolbox */</p>
<hr />
<div>== Introduction ==<br />
<br />
he blindDACE toolbox is a versatile Matlab toolbox that implements the popular Gaussian Process based kriging surrogate models. Kriging is in particular popular for approximating (and optimizing) deterministic computer experiments. Given a dataset the toolbox automatically fits a kriging surrogate model to it. Afterwards the kriging surrogate can be fully exploited instead of the (probably more expensive) simulation code.<br />
<br />
The toolbox is aimed for solving complex applications (expensive simulation codes, physical experiments, ...) and for researching new kriging extensions and techniques.<br />
<br />
== Download ==<br />
<br />
See: [http://sumo.intec.ugent.be/?q=blindDACE download page]<br />
<br />
== Quick start guide ==<br />
<br />
'''IMPORTANT: Before the toolbox can be used you have to include the toolbox in Matlab's path. You can do this manually by running startup or if Matlab is started in the root toolbox directory then startup will be run automatically.'''<br />
<br />
<source lang="matlab"><br />
startup<br />
</source><br />
<br />
<br />
Now the toolbox is ready to be used. The blindDACE toolbox is designed in an object oriented (OO) fashion.<br />
It is strongly recommended to exploit the OO design directly, i.e., use the Kriging and Optimizer matlab classes.<br />
However, for convenience wrapper scripts (dacefit, predictor) are provided that simulate the DACE toolbox interface (see [[#DACE toolbox interface|wrapper scripts]] for more information).<br />
<br />
<source lang="matlab"><br />
...<br />
% Generate kriging options structure<br />
opts = getDefaultOptions();<br />
opts.hpBounds = [lb ; ub]; % hyperparameter optimization bounds<br />
<br />
% configure the optimization algorithm (only one optimizer is included)<br />
% the Matlab Optimization toolbox is REQUIRED<br />
optimopts.GradObj = 'on';<br />
optimopts.DerivativeCheck = 'off';<br />
optimopts.Diagnostics = 'off';<br />
optimopts.Algorithm = 'active-set';<br />
opts.hpOptimizer = MatlabOptimizer( dim, 1, optimopts );<br />
<br />
% build and fit Kriging object<br />
k = Kriging( opts, theta0, 'regpoly0', @corrgauss );<br />
k = k.fit( samples, values );<br />
<br />
% k represents the approximation and can now be used, e.g.,<br />
[y mse] = k.predict( [1 2] )<br />
...<br />
</source><br />
<br />
See the included demo.m script for more example code on how to use the blindDACE toolbox (including more advances features such as using blind kriging or how to use regression instead of interpolation).<br />
<br />
== DACE toolbox interface ==<br />
<br />
The blindDACE toolbox provides two scripts dacefit.m and predictor.m that simulate the behavior of the DACE toolbox ([http://www2.imm.dtu.dk/~hbn/dace/]). Note, that full compatibility between blindDACE and the DACE toolbox is not provided. The scripts merely aim to ease the transition from the DACE toolbox to blindDACE.<br />
<br />
Example code:<br />
<source lang="matlab"><br />
krige = dacefit(samples, values, 'regpoly0', 'corrgauss', theta0, lb, ub )<br />
y = predictor([1 2], krige)<br />
</source><br />
<br />
Obviously, a lot less code is used to copy the setup described above. However, less code means less flexibility (e.g., blind kriging and regression kriging is not available using the wrapper scripts). Hence, it is suggested to learn the object oriented interface of blindDACE and use that instead.<br />
<br />
== Contribute ==<br />
<br />
These bindings are very basic but they work for me. Please improve, extend, provide binaries, and of course contribute back.</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=OoDACE:ooDACE_toolbox&diff=5072OoDACE:ooDACE toolbox2010-02-09T10:12:52Z<p>Icouckuy: /* Download */</p>
<hr />
<div>== blindDACE: A versatile kriging Matlab Toolbox ==<br />
<br />
TODO description<br />
<br />
== Download ==<br />
<br />
See: [http://sumo.intec.ugent.be/?q=blindDACE download page]<br />
<br />
== Quick start guide ==<br />
<br />
'''IMPORTANT: Before the toolbox can be used you have to include the toolbox in Matlab's path. You can do this manually by running startup or if Matlab is started in the root toolbox directory then startup will be run automatically.'''<br />
<br />
<source lang="matlab"><br />
startup<br />
</source><br />
<br />
<br />
Now the toolbox is ready to be used. The blindDACE toolbox is designed in an object oriented (OO) fashion.<br />
It is strongly recommended to exploit the OO design directly, i.e., use the Kriging and Optimizer matlab classes.<br />
However, for convenience wrapper scripts (dacefit, predictor) are provided that simulate the DACE toolbox interface (see [[#DACE toolbox interface|wrapper scripts]] for more information).<br />
<br />
<source lang="matlab"><br />
...<br />
% Generate kriging options structure<br />
opts = getDefaultOptions();<br />
opts.hpBounds = [lb ; ub]; % hyperparameter optimization bounds<br />
<br />
% configure the optimization algorithm (only one optimizer is included)<br />
% the Matlab Optimization toolbox is REQUIRED<br />
optimopts.GradObj = 'on';<br />
optimopts.DerivativeCheck = 'off';<br />
optimopts.Diagnostics = 'off';<br />
optimopts.Algorithm = 'active-set';<br />
opts.hpOptimizer = MatlabOptimizer( dim, 1, optimopts );<br />
<br />
% build and fit Kriging object<br />
k = Kriging( opts, theta0, 'regpoly0', @corrgauss );<br />
k = k.fit( samples, values );<br />
<br />
% k represents the approximation and can now be used, e.g.,<br />
[y mse] = k.predict( [1 2] )<br />
...<br />
</source><br />
<br />
See the included demo.m script for more example code on how to use the blindDACE toolbox (including more advances features such as using blind kriging or how to use regression instead of interpolation).<br />
<br />
== DACE toolbox interface ==<br />
<br />
The blindDACE toolbox provides two scripts dacefit.m and predictor.m that simulate the behavior of the DACE toolbox ([http://www2.imm.dtu.dk/~hbn/dace/]). Note, that full compatibility between blindDACE and the DACE toolbox is not provided. The scripts merely aim to ease the transition from the DACE toolbox to blindDACE.<br />
<br />
Example code:<br />
<source lang="matlab"><br />
krige = dacefit(samples, values, 'regpoly0', 'corrgauss', theta0, lb, ub )<br />
y = predictor([1 2], krige)<br />
</source><br />
<br />
Obviously, a lot less code is used to copy the setup described above. However, less code means less flexibility (e.g., blind kriging and regression kriging is not available using the wrapper scripts). Hence, it is suggested to learn the object oriented interface of blindDACE and use that instead.<br />
<br />
== Contribute ==<br />
<br />
These bindings are very basic but they work for me. Please improve, extend, provide binaries, and of course contribute back.</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=OoDACE:ooDACE_toolbox&diff=5071OoDACE:ooDACE toolbox2010-02-09T10:12:34Z<p>Icouckuy: /* Download */</p>
<hr />
<div>== blindDACE: A versatile kriging Matlab Toolbox ==<br />
<br />
TODO description<br />
<br />
== Download ==<br />
<br />
See: [[http://sumo.intec.ugent.be/?q=blindDACE|download page]]<br />
<br />
== Quick start guide ==<br />
<br />
'''IMPORTANT: Before the toolbox can be used you have to include the toolbox in Matlab's path. You can do this manually by running startup or if Matlab is started in the root toolbox directory then startup will be run automatically.'''<br />
<br />
<source lang="matlab"><br />
startup<br />
</source><br />
<br />
<br />
Now the toolbox is ready to be used. The blindDACE toolbox is designed in an object oriented (OO) fashion.<br />
It is strongly recommended to exploit the OO design directly, i.e., use the Kriging and Optimizer matlab classes.<br />
However, for convenience wrapper scripts (dacefit, predictor) are provided that simulate the DACE toolbox interface (see [[#DACE toolbox interface|wrapper scripts]] for more information).<br />
<br />
<source lang="matlab"><br />
...<br />
% Generate kriging options structure<br />
opts = getDefaultOptions();<br />
opts.hpBounds = [lb ; ub]; % hyperparameter optimization bounds<br />
<br />
% configure the optimization algorithm (only one optimizer is included)<br />
% the Matlab Optimization toolbox is REQUIRED<br />
optimopts.GradObj = 'on';<br />
optimopts.DerivativeCheck = 'off';<br />
optimopts.Diagnostics = 'off';<br />
optimopts.Algorithm = 'active-set';<br />
opts.hpOptimizer = MatlabOptimizer( dim, 1, optimopts );<br />
<br />
% build and fit Kriging object<br />
k = Kriging( opts, theta0, 'regpoly0', @corrgauss );<br />
k = k.fit( samples, values );<br />
<br />
% k represents the approximation and can now be used, e.g.,<br />
[y mse] = k.predict( [1 2] )<br />
...<br />
</source><br />
<br />
See the included demo.m script for more example code on how to use the blindDACE toolbox (including more advances features such as using blind kriging or how to use regression instead of interpolation).<br />
<br />
== DACE toolbox interface ==<br />
<br />
The blindDACE toolbox provides two scripts dacefit.m and predictor.m that simulate the behavior of the DACE toolbox ([http://www2.imm.dtu.dk/~hbn/dace/]). Note, that full compatibility between blindDACE and the DACE toolbox is not provided. The scripts merely aim to ease the transition from the DACE toolbox to blindDACE.<br />
<br />
Example code:<br />
<source lang="matlab"><br />
krige = dacefit(samples, values, 'regpoly0', 'corrgauss', theta0, lb, ub )<br />
y = predictor([1 2], krige)<br />
</source><br />
<br />
Obviously, a lot less code is used to copy the setup described above. However, less code means less flexibility (e.g., blind kriging and regression kriging is not available using the wrapper scripts). Hence, it is suggested to learn the object oriented interface of blindDACE and use that instead.<br />
<br />
== Contribute ==<br />
<br />
These bindings are very basic but they work for me. Please improve, extend, provide binaries, and of course contribute back.</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=OoDACE:ooDACE_toolbox&diff=5070OoDACE:ooDACE toolbox2010-02-09T10:12:13Z<p>Icouckuy: /* Download */</p>
<hr />
<div>== blindDACE: A versatile kriging Matlab Toolbox ==<br />
<br />
TODO description<br />
<br />
== Download ==<br />
<br />
See: [http://sumo.intec.ugent.be/?q=blindDACE|download page]]<br />
<br />
== Quick start guide ==<br />
<br />
'''IMPORTANT: Before the toolbox can be used you have to include the toolbox in Matlab's path. You can do this manually by running startup or if Matlab is started in the root toolbox directory then startup will be run automatically.'''<br />
<br />
<source lang="matlab"><br />
startup<br />
</source><br />
<br />
<br />
Now the toolbox is ready to be used. The blindDACE toolbox is designed in an object oriented (OO) fashion.<br />
It is strongly recommended to exploit the OO design directly, i.e., use the Kriging and Optimizer matlab classes.<br />
However, for convenience wrapper scripts (dacefit, predictor) are provided that simulate the DACE toolbox interface (see [[#DACE toolbox interface|wrapper scripts]] for more information).<br />
<br />
<source lang="matlab"><br />
...<br />
% Generate kriging options structure<br />
opts = getDefaultOptions();<br />
opts.hpBounds = [lb ; ub]; % hyperparameter optimization bounds<br />
<br />
% configure the optimization algorithm (only one optimizer is included)<br />
% the Matlab Optimization toolbox is REQUIRED<br />
optimopts.GradObj = 'on';<br />
optimopts.DerivativeCheck = 'off';<br />
optimopts.Diagnostics = 'off';<br />
optimopts.Algorithm = 'active-set';<br />
opts.hpOptimizer = MatlabOptimizer( dim, 1, optimopts );<br />
<br />
% build and fit Kriging object<br />
k = Kriging( opts, theta0, 'regpoly0', @corrgauss );<br />
k = k.fit( samples, values );<br />
<br />
% k represents the approximation and can now be used, e.g.,<br />
[y mse] = k.predict( [1 2] )<br />
...<br />
</source><br />
<br />
See the included demo.m script for more example code on how to use the blindDACE toolbox (including more advances features such as using blind kriging or how to use regression instead of interpolation).<br />
<br />
== DACE toolbox interface ==<br />
<br />
The blindDACE toolbox provides two scripts dacefit.m and predictor.m that simulate the behavior of the DACE toolbox ([http://www2.imm.dtu.dk/~hbn/dace/]). Note, that full compatibility between blindDACE and the DACE toolbox is not provided. The scripts merely aim to ease the transition from the DACE toolbox to blindDACE.<br />
<br />
Example code:<br />
<source lang="matlab"><br />
krige = dacefit(samples, values, 'regpoly0', 'corrgauss', theta0, lb, ub )<br />
y = predictor([1 2], krige)<br />
</source><br />
<br />
Obviously, a lot less code is used to copy the setup described above. However, less code means less flexibility (e.g., blind kriging and regression kriging is not available using the wrapper scripts). Hence, it is suggested to learn the object oriented interface of blindDACE and use that instead.<br />
<br />
== Contribute ==<br />
<br />
These bindings are very basic but they work for me. Please improve, extend, provide binaries, and of course contribute back.</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=OoDACE:ooDACE_toolbox&diff=5069OoDACE:ooDACE toolbox2010-02-09T10:10:52Z<p>Icouckuy: /* Quick start guide */</p>
<hr />
<div>== blindDACE: A versatile kriging Matlab Toolbox ==<br />
<br />
TODO description<br />
<br />
== Download ==<br />
<br />
See: [[http://sumo.intec.be/blindDACE]]<br />
<br />
== Quick start guide ==<br />
<br />
'''IMPORTANT: Before the toolbox can be used you have to include the toolbox in Matlab's path. You can do this manually by running startup or if Matlab is started in the root toolbox directory then startup will be run automatically.'''<br />
<br />
<source lang="matlab"><br />
startup<br />
</source><br />
<br />
<br />
Now the toolbox is ready to be used. The blindDACE toolbox is designed in an object oriented (OO) fashion.<br />
It is strongly recommended to exploit the OO design directly, i.e., use the Kriging and Optimizer matlab classes.<br />
However, for convenience wrapper scripts (dacefit, predictor) are provided that simulate the DACE toolbox interface (see [[#DACE toolbox interface|wrapper scripts]] for more information).<br />
<br />
<source lang="matlab"><br />
...<br />
% Generate kriging options structure<br />
opts = getDefaultOptions();<br />
opts.hpBounds = [lb ; ub]; % hyperparameter optimization bounds<br />
<br />
% configure the optimization algorithm (only one optimizer is included)<br />
% the Matlab Optimization toolbox is REQUIRED<br />
optimopts.GradObj = 'on';<br />
optimopts.DerivativeCheck = 'off';<br />
optimopts.Diagnostics = 'off';<br />
optimopts.Algorithm = 'active-set';<br />
opts.hpOptimizer = MatlabOptimizer( dim, 1, optimopts );<br />
<br />
% build and fit Kriging object<br />
k = Kriging( opts, theta0, 'regpoly0', @corrgauss );<br />
k = k.fit( samples, values );<br />
<br />
% k represents the approximation and can now be used, e.g.,<br />
[y mse] = k.predict( [1 2] )<br />
...<br />
</source><br />
<br />
See the included demo.m script for more example code on how to use the blindDACE toolbox (including more advances features such as using blind kriging or how to use regression instead of interpolation).<br />
<br />
== DACE toolbox interface ==<br />
<br />
The blindDACE toolbox provides two scripts dacefit.m and predictor.m that simulate the behavior of the DACE toolbox ([http://www2.imm.dtu.dk/~hbn/dace/]). Note, that full compatibility between blindDACE and the DACE toolbox is not provided. The scripts merely aim to ease the transition from the DACE toolbox to blindDACE.<br />
<br />
Example code:<br />
<source lang="matlab"><br />
krige = dacefit(samples, values, 'regpoly0', 'corrgauss', theta0, lb, ub )<br />
y = predictor([1 2], krige)<br />
</source><br />
<br />
Obviously, a lot less code is used to copy the setup described above. However, less code means less flexibility (e.g., blind kriging and regression kriging is not available using the wrapper scripts). Hence, it is suggested to learn the object oriented interface of blindDACE and use that instead.<br />
<br />
== Contribute ==<br />
<br />
These bindings are very basic but they work for me. Please improve, extend, provide binaries, and of course contribute back.</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=OoDACE:ooDACE_toolbox&diff=5068OoDACE:ooDACE toolbox2010-02-09T10:08:48Z<p>Icouckuy: /* DACE toolbox interface */</p>
<hr />
<div>== blindDACE: A versatile kriging Matlab Toolbox ==<br />
<br />
TODO description<br />
<br />
== Download ==<br />
<br />
See: [[http://sumo.intec.be/blindDACE]]<br />
<br />
== Quick start guide ==<br />
<br />
'''IMPORTANT: Before the toolbox can be used you have to include the toolbox in Matlab's path. You can do this manually by running startup or if Matlab is started in the root toolbox directory then startup will be run automatically.'''<br />
<br />
<source lang="matlab"><br />
startup<br />
</source><br />
<br />
<br />
Now the toolbox is ready to be used. The blindDACE toolbox is designed in an object oriented (OO) fashion.<br />
It is strongly recommended to exploit the OO design directly, i.e., use the Kriging and Optimizer matlab classes.<br />
However, for convenience wrapper scripts (dacefit, predictor) are provided that simulate the DACE toolbox interface (see section TODO for more information).<br />
<br />
<source lang="matlab"><br />
...<br />
% Generate kriging options structure<br />
opts = getDefaultOptions();<br />
opts.hpBounds = [lb ; ub]; % hyperparameter optimization bounds<br />
<br />
% configure the optimization algorithm (only one optimizer is included)<br />
% the Matlab Optimization toolbox is REQUIRED<br />
optimopts.GradObj = 'on';<br />
optimopts.DerivativeCheck = 'off';<br />
optimopts.Diagnostics = 'off';<br />
optimopts.Algorithm = 'active-set';<br />
opts.hpOptimizer = MatlabOptimizer( dim, 1, optimopts );<br />
<br />
% build and fit Kriging object<br />
k = Kriging( opts, theta0, 'regpoly0', @corrgauss );<br />
k = k.fit( samples, values );<br />
<br />
% k represents the approximation and can now be used, e.g.,<br />
[y mse] = k.predict( [1 2] )<br />
...<br />
</source><br />
<br />
See the included demo.m script for more example code on how to use the blindDACE toolbox (including more advances features such as using blind kriging or how to use regression instead of interpolation).<br />
<br />
== DACE toolbox interface ==<br />
<br />
The blindDACE toolbox provides two scripts dacefit.m and predictor.m that simulate the behavior of the DACE toolbox ([http://www2.imm.dtu.dk/~hbn/dace/]). Note, that full compatibility between blindDACE and the DACE toolbox is not provided. The scripts merely aim to ease the transition from the DACE toolbox to blindDACE.<br />
<br />
Example code:<br />
<source lang="matlab"><br />
krige = dacefit(samples, values, 'regpoly0', 'corrgauss', theta0, lb, ub )<br />
y = predictor([1 2], krige)<br />
</source><br />
<br />
Obviously, a lot less code is used to copy the setup described above. However, less code means less flexibility (e.g., blind kriging and regression kriging is not available using the wrapper scripts). Hence, it is suggested to learn the object oriented interface of blindDACE and use that instead.<br />
<br />
== Contribute ==<br />
<br />
These bindings are very basic but they work for me. Please improve, extend, provide binaries, and of course contribute back.</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=OoDACE:ooDACE_toolbox&diff=5067OoDACE:ooDACE toolbox2010-02-09T10:08:03Z<p>Icouckuy: /* DACE toolbox interface */</p>
<hr />
<div>== blindDACE: A versatile kriging Matlab Toolbox ==<br />
<br />
TODO description<br />
<br />
== Download ==<br />
<br />
See: [[http://sumo.intec.be/blindDACE]]<br />
<br />
== Quick start guide ==<br />
<br />
'''IMPORTANT: Before the toolbox can be used you have to include the toolbox in Matlab's path. You can do this manually by running startup or if Matlab is started in the root toolbox directory then startup will be run automatically.'''<br />
<br />
<source lang="matlab"><br />
startup<br />
</source><br />
<br />
<br />
Now the toolbox is ready to be used. The blindDACE toolbox is designed in an object oriented (OO) fashion.<br />
It is strongly recommended to exploit the OO design directly, i.e., use the Kriging and Optimizer matlab classes.<br />
However, for convenience wrapper scripts (dacefit, predictor) are provided that simulate the DACE toolbox interface (see section TODO for more information).<br />
<br />
<source lang="matlab"><br />
...<br />
% Generate kriging options structure<br />
opts = getDefaultOptions();<br />
opts.hpBounds = [lb ; ub]; % hyperparameter optimization bounds<br />
<br />
% configure the optimization algorithm (only one optimizer is included)<br />
% the Matlab Optimization toolbox is REQUIRED<br />
optimopts.GradObj = 'on';<br />
optimopts.DerivativeCheck = 'off';<br />
optimopts.Diagnostics = 'off';<br />
optimopts.Algorithm = 'active-set';<br />
opts.hpOptimizer = MatlabOptimizer( dim, 1, optimopts );<br />
<br />
% build and fit Kriging object<br />
k = Kriging( opts, theta0, 'regpoly0', @corrgauss );<br />
k = k.fit( samples, values );<br />
<br />
% k represents the approximation and can now be used, e.g.,<br />
[y mse] = k.predict( [1 2] )<br />
...<br />
</source><br />
<br />
See the included demo.m script for more example code on how to use the blindDACE toolbox (including more advances features such as using blind kriging or how to use regression instead of interpolation).<br />
<br />
== DACE toolbox interface ==<br />
<br />
The blindDACE toolbox provide two scripts dacefit.m and predictor.m that simulate the behavior of the DACE toolbox ([[http://www2.imm.dtu.dk/~hbn/dace/]]). Note, that full compatibility is not provided. The scripts only aim to ease the transition from the DACE toolbox to blindDACE.<br />
<br />
Example code:<br />
<source lang="matlab"><br />
krige = dacefit(samples, values, 'regpoly0', 'corrgauss', theta0, lb, ub )<br />
y = predictor([1 2], krige)<br />
</source><br />
<br />
Obviously, a lot less code is used to copy the setup described above. However, less code means less flexibility (e.g., blind kriging and regression kriging is not available using the wrapper scripts). Hence, it is suggested to learn the object oriented interface of blindDACE and use that instead.<br />
<br />
== Contribute ==<br />
<br />
These bindings are very basic but they work for me. Please improve, extend, provide binaries, and of course contribute back.</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=OoDACE:ooDACE_toolbox&diff=5066OoDACE:ooDACE toolbox2010-02-09T10:02:33Z<p>Icouckuy: /* Quick start guide */</p>
<hr />
<div>== blindDACE: A versatile kriging Matlab Toolbox ==<br />
<br />
TODO description<br />
<br />
== Download ==<br />
<br />
See: [[http://sumo.intec.be/blindDACE]]<br />
<br />
== Quick start guide ==<br />
<br />
'''IMPORTANT: Before the toolbox can be used you have to include the toolbox in Matlab's path. You can do this manually by running startup or if Matlab is started in the root toolbox directory then startup will be run automatically.'''<br />
<br />
<source lang="matlab"><br />
startup<br />
</source><br />
<br />
<br />
Now the toolbox is ready to be used. The blindDACE toolbox is designed in an object oriented (OO) fashion.<br />
It is strongly recommended to exploit the OO design directly, i.e., use the Kriging and Optimizer matlab classes.<br />
However, for convenience wrapper scripts (dacefit, predictor) are provided that simulate the DACE toolbox interface (see section TODO for more information).<br />
<br />
<source lang="matlab"><br />
...<br />
% Generate kriging options structure<br />
opts = getDefaultOptions();<br />
opts.hpBounds = [lb ; ub]; % hyperparameter optimization bounds<br />
<br />
% configure the optimization algorithm (only one optimizer is included)<br />
% the Matlab Optimization toolbox is REQUIRED<br />
optimopts.GradObj = 'on';<br />
optimopts.DerivativeCheck = 'off';<br />
optimopts.Diagnostics = 'off';<br />
optimopts.Algorithm = 'active-set';<br />
opts.hpOptimizer = MatlabOptimizer( dim, 1, optimopts );<br />
<br />
% build and fit Kriging object<br />
k = Kriging( opts, theta0, 'regpoly0', @corrgauss );<br />
k = k.fit( samples, values );<br />
<br />
% k represents the approximation and can now be used, e.g.,<br />
[y mse] = k.predict( [1 2] )<br />
...<br />
</source><br />
<br />
See the included demo.m script for more example code on how to use the blindDACE toolbox (including more advances features such as using blind kriging or how to use regression instead of interpolation).<br />
<br />
== DACE toolbox interface ==<br />
<br />
<code><pre>dacefit()</pre></code><br />
<code><pre>predictor()</pre></code><br />
<br />
== Contribute ==<br />
<br />
These bindings are very basic but they work for me. Please improve, extend, provide binaries, and of course contribute back.</div>Icouckuyhttp://sumowiki.intec.ugent.be/index.php?title=OoDACE:ooDACE_toolbox&diff=5065OoDACE:ooDACE toolbox2010-02-09T10:00:46Z<p>Icouckuy: /* Quick start guide */</p>
<hr />
<div>== blindDACE: A versatile kriging Matlab Toolbox ==<br />
<br />
TODO description<br />
<br />
== Download ==<br />
<br />
See: [[http://sumo.intec.be/blindDACE]]<br />
<br />
== Quick start guide ==<br />
<br />
'''IMPORTANT: Before the toolbox can be used you have to include the toolbox in Matlab's path. You can do this manually by running startup or if Matlab is started in the root toolbox directory then startup will be run automatically.'''<br />
<br />
<source lang="matlab"><br />
startup<br />
</source><br />
<br />
<br />
Now the toolbox is ready to be used. The blindDACE toolbox is designed in an object oriented (OO) fashion.<br />
It is strongly recommended to exploit the OO design directly, i.e., use the Kriging and Optimizer matlab classes.<br />
However, for convenience wrapper scripts (dacefit, predictor) are provided that simulate the DACE toolbox interface (see section TODO for more information).<br />
<br />
<source lang="matlab"><br />
...<br />
% Generate kriging options structure<br />
opts = getDefaultOptions();<br />
opts.hpBounds = [lb ; ub]; % hyperparameter optimization bounds<br />
<br />
% configure the optimization algorithm (only one optimizer is included)<br />
% the Matlab Optimization toolbox is REQUIRED<br />
optimopts.GradObj = 'on';<br />
optimopts.DerivativeCheck = 'off';<br />
optimopts.Diagnostics = 'off';<br />
optimopts.Algorithm = 'active-set';<br />
opts.hpOptimizer = MatlabOptimizer( dim, 1, optimopts );<br />
<br />
% required BEFORE constructing Kriging class<br />
switch(id)<br />
case 2<br />
% add regression parameter<br />
opts.lambda0 = 0;<br />
opts.lambdaBounds = [-10 ; 0]; % log scale<br />
case 3<br />
% enable blind kriging<br />
opts.regressionMetric = 'cvpe';<br />
end<br />
</source><br />
<br />
See the included demo.m script for more example code on how to use the blindDACE toolbox.<br />
<br />
== DACE toolbox interface ==<br />
<br />
<code><pre>dacefit()</pre></code><br />
<code><pre>predictor()</pre></code><br />
<br />
== Contribute ==<br />
<br />
These bindings are very basic but they work for me. Please improve, extend, provide binaries, and of course contribute back.</div>Icouckuy