CovariateShiftSplitter<T>

Models & Types ยท AiDotNet.Preprocessing.DataPreparation.Splitting.Specialized

Covariate shift splitter that creates intentional distribution shift between train and test.

Inheritance: DataSplitterBase<T> โ†’ CovariateShiftSplitter<T>

For Beginners

Covariate shift occurs when the distribution of input features differs between training and test data, even though the relationship between inputs and outputs remains the same. This splitter intentionally creates such a shift.

How It Works

How It Works:

  1. Identify a primary feature dimension for the shift
  2. Select training samples from one range of that feature
  3. Select test samples from a different range

When to Use:

  • Testing model robustness to distribution shift
  • Simulating real-world deployment scenarios
  • Evaluating domain adaptation techniques

Constructors

Constructor Summary
CovariateShiftSplitter(double testSize, double shiftStrength, int shiftDimension, int randomSeed) Creates a new covariate shift splitter.

Properties

Property Summary
string Description { get; }

Methods

Method Summary
DataSplitResult<T> Split(Matrix<T> X, Vector<T> y)