CorrelationRemovalSelector<T>

Models & Types ยท AiDotNet.Preprocessing.FeatureSelection.Unsupervised

Correlation-based Feature Removal.

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

Implements: IFeatureSelector<T>

For Beginners

When two features are highly correlated, they provide similar information. This method finds pairs of highly correlated features and removes one from each pair, keeping your dataset compact without losing much information.

How It Works

Removes highly correlated features to reduce multicollinearity, keeping one feature from each group of highly correlated features.

Constructors

Constructor Summary
CorrelationRemovalSelector(double threshold, int[] columnIndices)

Properties

Property Summary
Matrix<T> CorrelationMatrix { get; }
bool SupportsInverseTransform { get; }
double Threshold { get; }
int[] SelectedIndices { get; }

Methods

Method Summary
bool[] GetSupportMask()
string[] GetFeatureNamesOut(string[] inputFeatureNames)