ReliefMLC<T>

Models & Types ยท AiDotNet.Preprocessing.FeatureSelection.MultiLabel

Relief for Multi-Label Classification (ReliefMLC).

Inheritance: TransformerBase<T, Matrix<T>, Matrix<T>> โ†’ ReliefMLC<T>

Implements: IFixedCountFeatureSelector<T>

For Beginners

Standard Relief compares single labels to find hits/misses. ReliefMLC compares label sets, accounting for partial overlap in multi-label scenarios.

How It Works

Extends the Relief algorithm to handle multi-label problems. Uses Hamming distance on label sets to determine similarity between instances.

Constructors

Constructor Summary
ReliefMLC(int nFeaturesToSelect, int nNeighbors, int nIterations, int? randomState, int[] columnIndices)

Properties

Property Summary
Vector<T> FeatureWeights { get; }
bool SupportsInverseTransform { get; }
int NFeaturesToSelect { get; }
int[] SelectedIndices { get; }

Methods

Method Summary
Matrix<T> FitTransform(Matrix<T> data, Matrix<T> targets)
bool[] GetSupportMask()
string[] GetFeatureNamesOut(string[] inputFeatureNames)
void Fit(Matrix<T> data, Matrix<T> targets)