FastIcaWhiteningMode Enumeration

Specifies the whitening strategy applied to the input data before ICA extraction.

Definition

Namespace: Numerics.NET.Statistics.Multivariate
Assembly: Numerics.NET (in Numerics.NET.dll) Version: 9.1.5
C# 
public enum FastIcaWhiteningMode

Remarks

Whitening transforms the input data to have uncorrelated features and unit or arbitrary variance. The choice affects the scaling and interpretation of the extracted components.

  • Skip: Assumes data is already whitened. No whitening or dimensionality reduction is performed. NumberOfComponents must equal the number of features.
  • ArbitraryVariance: Applies whitening so that the resulting covariance is a scalar multiple of the identity matrix. Matches scikit-learn's 'arbitrary-variance' mode.
  • UnitVariance: Applies whitening and rescales the final sources to have unit variance. Matches scikit-learn's 'unit-variance' mode and is the default.

Members

Skip0 Skip whitening: assumes input data is already whitened. No transformation is performed.

NumberOfComponents must equal the number of features.

ArbitraryVariance1 Arbitrary variance whitening: transforms data so that the covariance matrix is a scalar multiple of the identity.

Matches scikit-learn's 'arbitrary-variance' option.

UnitVariance2 Unit variance whitening: transforms data and rescales sources to have unit variance.

Matches scikit-learn's 'unit-variance' option. This is the default.

See Also