LaplacianScore<T>

Models & Types ยท AiDotNet.Preprocessing.FeatureSelection.Filter.Spectral

Laplacian Score for unsupervised feature selection.

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

Implements: IFixedCountFeatureSelector<T>

For Beginners

The Laplacian Score measures how smoothly a feature varies across similar data points. A good feature should have similar values for nearby points (like neighbors having similar house prices). Features that jump around randomly between neighbors score poorly.

How It Works

Laplacian Score evaluates features based on their ability to preserve locality in the data. Features with low Laplacian Score preserve local structure well, meaning similar samples have similar feature values.

Constructors

Constructor Summary
LaplacianScore(int nFeaturesToSelect, int nNeighbors, double kernelWidth, int[] columnIndices)

Properties

Property Summary
Vector<T> LaplacianScores { get; }
bool SupportsInverseTransform { get; }
int NFeaturesToSelect { get; }
int NNeighbors { get; }
int[] SelectedIndices { get; }

Methods

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