DiverseEnsemble<T>

Models & Types ยท AiDotNet.Preprocessing.FeatureSelection.Ensemble

Diverse Ensemble Feature Selection using multiple diverse methods.

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

Implements: IFixedCountFeatureSelector<T>, ISupervisedFeatureSelector<T>

For Beginners

Different feature selection methods look at features from different angles - some focus on correlation, others on prediction power, others on regularization. By combining diverse perspectives, we identify features that are truly important from multiple viewpoints.

How It Works

Combines multiple feature selection methods with intentionally different characteristics (filter, wrapper, embedded) to get robust rankings. Features consistently ranked high across diverse methods are selected.

Constructors

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

Properties

Property Summary
Vector<T> AggregatedScores { 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)