StreamingFeatureSelector<T>

Models & Types · AiDotNet.Preprocessing.FeatureSelection.Incremental

Streaming Feature Selector for online/incremental learning.

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

Implements: IFixedCountFeatureSelector<T>, ISupervisedFeatureSelector<T>

For Beginners

When you can’t store all your data (it’s too big or comes continuously), you need to update feature scores incrementally. This method keeps running averages that update with each new sample, allowing feature selection on endless data streams.

How It Works

Maintains running statistics to update feature scores as new data arrives. This enables feature selection on streaming data without storing all historical samples.

Constructors

Constructor Summary
StreamingFeatureSelector(int nFeaturesToSelect, double decayFactor, int[] columnIndices)

Properties

Property Summary
Vector<T> CorrelationScores { get; }
bool SupportsInverseTransform { get; }
double DecayFactor { get; }
int NFeaturesToSelect { get; }
int SampleCount { 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)
void PartialFit(Matrix<T> data, Vector<T> target)
void Reset()