SemiSupervisedLaplacian<T>

Models & Types · AiDotNet.Preprocessing.FeatureSelection.SemiSupervised

Semi-Supervised Feature Selection using Laplacian regularization.

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

Implements: IFixedCountFeatureSelector<T>

For Beginners

When you have few labeled samples but many unlabeled ones, this method uses both. It builds a neighborhood graph from all data to understand the data’s structure, while using labels where available.

How It Works

Combines supervised information from labeled samples with the manifold structure captured from all samples (labeled and unlabeled) using the graph Laplacian.

Constructors

Constructor Summary
SemiSupervisedLaplacian(int nFeaturesToSelect, int nNeighbors, double alpha, int[] columnIndices)

Properties

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

Methods

Method Summary
Matrix<T> FitTransform(Matrix<T> data, Vector<T> target, bool[] isLabeled)
bool[] GetSupportMask()
string[] GetFeatureNamesOut(string[] inputFeatureNames)
void Fit(Matrix<T> data, Vector<T> target, bool[] isLabeled)