ai.libs.jaicore.ml.dyadranking.algorithm.featuretransform

Interface IDyadFeatureTransform

All Known Implementing Classes:

BiliniearFeatureTransform

public interface IDyadFeatureTransform

Feature transformation interface for the FeatureTransformPLDyadRanker.

Method Summary

Modifier and Type	Method and Description
default java.util.Map<IDyadRankingInstance,java.util.Map<Dyad,ai.libs.jaicore.math.linearalgebra.Vector>>	getPreComputedFeatureTransforms(DyadRankingDataset dataset) Precomputed the feature transforms for the dataset, this can speed up the runtime as the feature transform will be reduced to O(1) at the cost of O(n).
int	getTransformedVectorLength(int alternativeLength, int instanceLength) Get the length of the vector returned by the transform method.
ai.libs.jaicore.math.linearalgebra.Vector	transform(Dyad dyad) Transform the instance of the given dyad (models the skill).

Method Detail

transform

ai.libs.jaicore.math.linearalgebra.Vector transform(Dyad dyad)

Transform the instance of the given dyad (models the skill).

Parameters:

dyad - the dyad to transform

Returns:

the transformed instance values for the dyad

getTransformedVectorLength

int getTransformedVectorLength(int alternativeLength,
                               int instanceLength)

Get the length of the vector returned by the transform method.

Parameters:

alternativeLength - the length of the alternative vector of the transformed dyad

instanceLength - the length of the instance vector of the transformed dyad

Returns:

the length of the transformed feature vector

getPreComputedFeatureTransforms

default java.util.Map<IDyadRankingInstance,java.util.Map<Dyad,ai.libs.jaicore.math.linearalgebra.Vector>> getPreComputedFeatureTransforms(DyadRankingDataset dataset)

Precomputed the feature transforms for the dataset, this can speed up the runtime as the feature transform will be reduced to O(1) at the cost of O(n).

Parameters:

dataset -

Returns:

the feature transform