Synopsis
A cost evaluator delivers as output the costs for given classification results.
Description
This operator provides the ability to evaluate classification costs. Therefore a cost matrix might be specified, denoting the costs for every possible classification outcome: predicted label x real label. Costs will be minimized during optimization.
Input
- example set: expects: ExampleSet
Output
- example set:
- performance:
Parameters
- keep exampleSet: Indicates if the example set should be kept.
- cost matrix: The matrix of missclassification costs. Columns and Rows in order of internal mapping.