CorrelationBasedFS<T>

Models & Types · AiDotNet.Preprocessing.FeatureSelection.Filter

Correlation-based Feature Selection (CFS).

Inheritance: TransformerBase<T, Matrix<T>, Matrix<T>> → CorrelationBasedFS<T>

Implements: IFixedCountFeatureSelector<T>, ISupervisedFeatureSelector<T>

For Beginners

CFS looks for features that are strongly related to what you want to predict but not too similar to each other. Having two features that are almost identical doesn’t help much - you want diverse information sources that all point toward the target.

How It Works

CFS evaluates feature subsets by considering both feature-target correlation (relevance) and feature-feature correlation (redundancy). Good subsets have high correlation with the target but low intercorrelation among features.

Constructors

Constructor Summary
CorrelationBasedFS(int nFeaturesToSelect, int[] columnIndices)

Properties

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

Methods

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