KernelFisherSelector<T>

Models & Types · AiDotNet.Preprocessing.FeatureSelection.Kernel

Kernel Fisher Discriminant-based Feature Selection.

Inheritance: TransformerBase<T, Matrix<T>, Matrix<T>> → KernelFisherSelector<T>

Implements: IFixedCountFeatureSelector<T>, ISupervisedFeatureSelector<T>

For Beginners

Fisher’s criterion tries to maximize the distance between class means while minimizing the spread within each class. Doing this in kernel space allows us to find non-linear boundaries between classes, selecting features that help separate classes in complex ways.

How It Works

Applies the Fisher criterion in kernel space to find features that maximize class separation using non-linear transformations.

Constructors

Constructor Summary
KernelFisherSelector(int nFeaturesToSelect, double gamma, int[] columnIndices)

Properties

Property Summary
Vector<T> FisherScores { get; }
bool SupportsInverseTransform { get; }
double Gamma { 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)