Package ai.libs.jaicore.ml.dyadranking.loss

Class DyadRankingLossUtil

java.lang.Object

ai.libs.jaicore.ml.dyadranking.loss.DyadRankingLossUtil

public class DyadRankingLossUtil
extends java.lang.Object

Class that contains utility methods for handling dyad ranking losses.

Method Summary

Modifier and Type	Method	Description
static double	computeAverageLoss(DyadRankingLossFunction lossFunction, DyadRankingDataset trueOrderings, IDyadRanker ranker)
static double	computeAverageLoss(DyadRankingLossFunction lossFunction, DyadRankingDataset trueOrderings, IDyadRanker ranker, java.util.Random random)	Computes the average loss over several dyad orderings.
static double	computeAverageLoss(DyadRankingLossFunction lossFunction, DyadRankingDataset trueOrderings, DyadRankingDataset predictedOrderings)	Computes the average loss over several dyad orderings.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Method Detail

computeAverageLoss

public static double computeAverageLoss(DyadRankingLossFunction lossFunction,
                                        DyadRankingDataset trueOrderings,
                                        DyadRankingDataset predictedOrderings)

Computes the average loss over several dyad orderings.

Parameters:

lossFunction - The loss function to be used for the individual IDyadRankingInstances

trueOrderings - The true orderings represented by IDyadRankingInstances

predictedOrderings - The predicted orderings represented by IDyadRankingInstances

Returns:

Average loss over all IDyadRankingInstances

computeAverageLoss

public static double computeAverageLoss(DyadRankingLossFunction lossFunction,
                                        DyadRankingDataset trueOrderings,
                                        IDyadRanker ranker,
                                        java.util.Random random)
                                 throws PredictionException

Computes the average loss over several dyad orderings. Predictions are obtained by the given IDyadRanker.

Parameters:

lossFunction - The loss function to be used for the individual IDyadRankingInstances

trueOrderings - The true orderings represented by IDyadRankingInstances

ranker - The IDyadRanker used to make predictions

Returns:

Average loss over all IDyadRankingInstances

Throws:

PredictionException

computeAverageLoss

public static double computeAverageLoss(DyadRankingLossFunction lossFunction,
                                        DyadRankingDataset trueOrderings,
                                        IDyadRanker ranker)
                                 throws PredictionException

Throws:

PredictionException