Previous Table of Contents Next


12.3.31 ClassificationFunctionSettings


   A ClassificationFunctionSettings object is a subclass of SupervisedFunctionSettings that supports features unique to the classification mining function and corresponding algorithms, specifically CostMatrix. The CostMatrix must be associated with the target LogicalAttribute.

   Superclasses

   SupervisedFunctionSettings

   References

   costMatrix

   The optional CostMatrix attribute specifies a two-dimensional, NxN matrix that defines the cost associated with a prediction versus the actual value. A cost matrix is typically used in classification models, where N is the number of classes in the target, and the columns and rows are labeled with class values.

class: CategoryMatrix
defined by: ClassificationSettingsRefCostMatrix
multiplicity: zero or one
inverse: CategoryMatrix::settings