AutoencoderSelector<T>

Models & Types ยท AiDotNet.Preprocessing.FeatureSelection.Neural

Autoencoder based Feature Selection.

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

Implements: IFixedCountFeatureSelector<T>

For Beginners

An autoencoder compresses data and reconstructs it. Features that are harder to reconstruct accurately are often more important because they carry unique information not captured by other features.

How It Works

Selects features based on their reconstruction importance in a simple autoencoder, measuring how much each feature contributes to the learned representation.

Constructors

Constructor Summary
AutoencoderSelector(int nFeaturesToSelect, int hiddenSize, int epochs, double learningRate, int? randomState, int[] columnIndices)

Properties

Property Summary
Vector<T> ReconstructionImportance { get; }
bool SupportsInverseTransform { get; }
double LearningRate { get; }
int Epochs { get; }
int HiddenSize { get; }
int NFeaturesToSelect { get; }
int[] SelectedIndices { get; }

Methods

Method Summary
Matrix<T> FitTransform(Matrix<T> data)
bool[] GetSupportMask()
string[] GetFeatureNamesOut(string[] inputFeatureNames)