BoostingFeatureSelector<T>

Models & Types ยท AiDotNet.Preprocessing.FeatureSelection.Ensemble

Boosting-based Feature Selection.

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

Implements: IFixedCountFeatureSelector<T>, ISupervisedFeatureSelector<T>

For Beginners

Like gradient boosting, this method focuses more attention on the samples that are hardest to predict. Features that help predict these difficult samples get higher importance scores, leading to more discriminative feature selection.

How It Works

Uses a boosting-inspired approach where samples are reweighted based on prediction errors, focusing feature selection on harder-to-predict samples.

Constructors

Constructor Summary
BoostingFeatureSelector(int nFeaturesToSelect, int nRounds, double learningRate, int[] columnIndices)

Properties

Property Summary
Vector<T> BoostingScores { get; }
bool SupportsInverseTransform { get; }
int NFeaturesToSelect { get; }
int NRounds { 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)