VariationalAutoencoderSelector<T>

Models & Types ยท AiDotNet.Preprocessing.FeatureSelection.Neural

Variational Autoencoder-based Feature Selection.

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

Implements: IFixedCountFeatureSelector<T>

For Beginners

A VAE learns not just a compressed representation but a probability distribution over possible representations. Features that contribute most to this learned distribution are the important ones.

How It Works

Uses a Variational Autoencoder (VAE) to learn a probabilistic latent space, then selects features based on their contribution to the learned distribution.

Constructors

Constructor Summary
VariationalAutoencoderSelector(int nFeaturesToSelect, int latentDim, int nEpochs, int[] columnIndices)

Properties

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

Methods

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