VarianceThreshold<T>

Models & Types · AiDotNet.Preprocessing.FeatureSelection.Filter.Univariate

Feature selector that removes features with variance below a threshold.

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

For Beginners

If a feature has the same value (or nearly the same value) for all samples, it won’t help your model distinguish between different outcomes. This transformer automatically removes such features:

  • Constant features (all same value) have variance = 0
  • Near-constant features have very low variance

Example: A “Country” column that is “USA” for 99.9% of rows provides little information.

How It Works

VarianceThreshold removes all features whose variance doesn’t meet a minimum threshold. Features with low variance are often not informative because they don’t vary enough across samples to be useful for prediction.

Constructors

Constructor Summary
VarianceThreshold(double threshold, int[] columnIndices) Creates a new instance of VarianceThreshold.

Properties

Property Summary
Vector<T> Variances { get; } Gets the computed variances for each feature.
bool SupportsInverseTransform { get; } Gets whether this transformer supports inverse transformation.
double Threshold { get; } Gets the variance threshold.
int[] SelectedFeatures { get; } Gets the indices of selected features.

Methods

Method Summary
bool[] GetSupportMask() Gets a boolean mask indicating which features are selected.
string[] GetFeatureNamesOut(string[] inputFeatureNames) Gets the output feature names after transformation.