GradientBoostingImportance<T>

Models & Types ยท AiDotNet.Preprocessing.FeatureSelection.Embedded

Gradient Boosting-based feature importance using iterative boosting.

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

Implements: IFixedCountFeatureSelector<T>, ISupervisedFeatureSelector<T>

For Beginners

Gradient Boosting builds trees sequentially, where each tree tries to fix the errors of previous trees. The features that help the most in reducing these errors across all trees are considered important.

How It Works

Gradient Boosting importance measures how often each feature is used for splits and how much those splits contribute to model improvement across all boosting rounds. Features that are frequently used for impactful splits score higher.

Constructors

Constructor Summary
GradientBoostingImportance(int nFeaturesToSelect, int nEstimators, int maxDepth, double learningRate, int? randomState, int[] columnIndices)

Properties

Property Summary
Vector<T> Importances { get; }
bool SupportsInverseTransform { get; }
double LearningRate { get; }
int NEstimators { 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)