FloatingForwardSelector<T>

Models & Types · AiDotNet.Preprocessing.FeatureSelection.Sequential

Sequential Floating Forward Selection (SFFS).

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

Implements: IFixedCountFeatureSelector<T>, ISupervisedFeatureSelector<T>

For Beginners

Regular forward selection only adds features but never removes them. SFFS adds features but can also remove them if doing so improves the result. This “floating” behavior helps find better feature combinations.

How It Works

Combines forward selection with conditional backward steps to escape local optima, allowing previously added features to be removed if they become redundant.

Constructors

Constructor Summary
FloatingForwardSelector(int nFeaturesToSelect, int maxBacktracks, int[] columnIndices)

Properties

Property Summary
Vector<T> FeatureImportances { 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)