Difference between revisions of "Measures"
| Line 28: | Line 28: | ||
{{Option | {{Option | ||
|name = modelDifference | |name = modelDifference | ||
| − | |values = | + | |values = [0,1] |
|default = 0 | |default = 0 | ||
|description = When this option is turned on, the average is calculated of the models built in different folds, and the score is based partly on the cross validation error and partly on the error of the average model in the samples. This option can be used to detect poles that will otherwise go unnoticed using the CrossValidation measure. A sensible value is 0.2. | |description = When this option is turned on, the average is calculated of the models built in different folds, and the score is based partly on the cross validation error and partly on the error of the average model in the samples. This option can be used to detect poles that will otherwise go unnoticed using the CrossValidation measure. A sensible value is 0.2. | ||
Revision as of 13:47, 27 June 2007
A measure is used to measure the quality of a new model. The measure decides wether the modelling process can be halted (the target accuracy has been reached) or wether new samples should be selected / new models should be built. The choice of measure is therefore very important for the success of the toolbox.
The rule of thumb is that the default measure, CrossValidation, is the best choice, but is also very expensive, as it requires that a number (5 by default) of models are built for each new model that has to be evaluated. Especially when using neural networks this can become an unacceptable overhead. If the modelling takes unacceptably long, the best alternative is TestSamples, which by default behaves as cross validation in which only one fold is considered, reducing the cost of the measure by a factor 5.
However, in certain situations it might be very effective to use different measures, or use multiple measures together. When multiple measures are used, an intelligent pareto-based method is used to decide which model is the best choice. Models that score high on a particular measure but low on another are not discarded immediately, but are given a chance to set things right in further iterations of the toolbox. This encourages variety in the models, while still ensuring convergence to the optimal accuracy for each measure. An often used combination is CrossValidation with the MinMax measure, to ensure that no poles are present in the model domain.
Below is a list of available measures and the configuration options available for each of them. Each measure also has a target accuracy attribute, which can be ommitted and which defaults to 0.001. In certain cases, such as the binary MinMax measure, the target accuracy is irrelevant.
CrossValidation
The CrossValidation measure is the default choice and performs an n-fold cross validation on the model to create an efficient estimation of the accuracy of the model. Several options are available to customize this measure.
Template:OptionsHeader Template:Option Template:Option Template:Option
TestSamples
LeaveNOut
MinMax
the MinMax measure is used to eliminate models that go below a given minimum or above a given maximum. This measure can be used to detect models that have poles in the model domain and to guide the modelling process in the right direction. If the output is known to lie within certain value bounds, these can be added to the simulator file as follows:
<OutputParameters> <Parameter name="out" type="real" minimum="-1" maximum="1"/> </OutputParameters>
When the MinMax measure is defined, these values will be used to ensure that all models stay within these bounds. If only the minimum or only the maximum is defined, naturally only these are enforced. There are no further configuration options for this measure.
