- Type Parameters:
InputType - Inputs of the Evaluator, such as a VectorTargetType - Targets of the Evaluator, such a Double
- All Superinterfaces:
- java.lang.Cloneable, CloneableSerializable, CostFunction<Evaluator<? super InputType,? extends TargetType>,java.util.Collection<? extends InputOutputPair<? extends InputType,TargetType>>>, Evaluator<Evaluator<? super InputType,? extends TargetType>,java.lang.Double>, PerformanceEvaluator<Evaluator<? super InputType,? extends TargetType>,java.util.Collection<? extends InputOutputPair<? extends InputType,TargetType>>,java.lang.Double>, java.io.Serializable, Summarizer<TargetEstimatePair<? extends TargetType,? extends TargetType>,java.lang.Double>, SupervisedPerformanceEvaluator<InputType,TargetType,TargetType,java.lang.Double>
- All Known Subinterfaces:
- DifferentiableCostFunction, ParallelizableCostFunction
- All Known Implementing Classes:
- AbstractParallelizableCostFunction, AbstractSupervisedCostFunction, MeanL1CostFunction, MeanSquaredErrorCostFunction, MeanZeroOneErrorEvaluator, ParallelizedCostFunctionContainer, SumSquaredErrorCostFunction
public interface SupervisedCostFunction<InputType,TargetType>
extends CostFunction<Evaluator<? super InputType,? extends TargetType>,java.util.Collection<? extends InputOutputPair<? extends InputType,TargetType>>>, Summarizer<TargetEstimatePair<? extends TargetType,? extends TargetType>,java.lang.Double>, SupervisedPerformanceEvaluator<InputType,TargetType,TargetType,java.lang.Double>
A type of CostFunction normally used in supervised-learning applications.
This CostFunction evaluates an Evaluator against its ability to reproduce
the targets from a collection of InputOutputPairs
- Since:
- 2.0
- Author:
- Kevin R. Dixon