ai.libs.jaicore.ml.dyadranking.optimizing

Class DyadRankingFeatureTransformNegativeLogLikelihoodDerivative

java.lang.Object

ai.libs.jaicore.ml.dyadranking.optimizing.DyadRankingFeatureTransformNegativeLogLikelihoodDerivative

All Implemented Interfaces:

IGradientFunction, IDyadRankingFeatureTransformPLGradientFunction

public class DyadRankingFeatureTransformNegativeLogLikelihoodDerivative
extends java.lang.Object
implements IDyadRankingFeatureTransformPLGradientFunction

Represents the derivate of the negative log likelihood function in the context of feature transformation Placket-Luce dyad ranking [1]. This implementation can be used for the partial derivatives of the linear vector w w.r.t. the negative log-likelihood that should be minimized. [1] Schäfer, D. & Hüllermeier, Dyad ranking using Plackett–Luce models based on joint feature representations, https://link.springer.com/article/10.1007%2Fs10994-017-5694-9

Constructor Summary

Constructors

Constructor and Description
DyadRankingFeatureTransformNegativeLogLikelihoodDerivative()

Method Summary

Modifier and Type	Method and Description
ai.libs.jaicore.math.linearalgebra.Vector	apply(ai.libs.jaicore.math.linearalgebra.Vector vector) Returns the result of applying the gradient to the point represented by the given vector.
void	initialize(DyadRankingDataset dataset, java.util.Map<IDyadRankingInstance,java.util.Map<Dyad,ai.libs.jaicore.math.linearalgebra.Vector>> featureTransforms) Initialize the function with the given data set and feature transformation method.

Methods inherited from class java.lang.Object

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

Constructor Detail

DyadRankingFeatureTransformNegativeLogLikelihoodDerivative

public DyadRankingFeatureTransformNegativeLogLikelihoodDerivative()

Method Detail

initialize

public void initialize(DyadRankingDataset dataset,
                       java.util.Map<IDyadRankingInstance,java.util.Map<Dyad,ai.libs.jaicore.math.linearalgebra.Vector>> featureTransforms)

Description copied from interface: IDyadRankingFeatureTransformPLGradientFunction

Initialize the function with the given data set and feature transformation method.

Specified by:

initialize in interface IDyadRankingFeatureTransformPLGradientFunction

Parameters:

dataset - the dataset to use

featureTransforms - the pre computed feature transformations

apply

public ai.libs.jaicore.math.linearalgebra.Vector apply(ai.libs.jaicore.math.linearalgebra.Vector vector)

Description copied from interface: IGradientFunction

Returns the result of applying the gradient to the point represented by the given vector.

Specified by:

apply in interface IGradientFunction

Parameters:

vector - the vector the gradient is applied to

Returns:

the new vector resulting from applying the gradient to the given vector