BidirectionalSearch<T>

Models & Types ยท AiDotNet.Preprocessing.FeatureSelection.Wrapper

Bidirectional Feature Search for feature selection.

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

Implements: IFixedCountFeatureSelector<T>, ISupervisedFeatureSelector<T>

For Beginners

Instead of just adding features (forward) or just removing them (backward), this method does both alternately. This can escape local optima and find feature combinations that neither approach would find alone.

How It Works

Bidirectional search combines forward selection and backward elimination in an interleaved manner. It can add or remove features at each step, potentially finding better solutions than either method alone.

Constructors

Constructor Summary
BidirectionalSearch(int nFeaturesToSelect, int maxIterations, int? randomState, int[] columnIndices)

Properties

Property Summary
Vector<T> FeatureScores { get; }
bool SupportsInverseTransform { get; }
int MaxIterations { 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)