Module: model_remediation.min_diff.losses

Stay organized with collections Save and categorize content based on your preferences.

View source on GitHub

API for MinDiff losses.

Modules

kernels module: API for MinDiff kernels.

Classes

class AbsoluteCorrelationLoss: Absolute correlation between predictions on two groups of examples.

class AdjustedMMDLoss: Adjusted Maximum Mean Discrepancy between predictions on two groups of examples.

class GaussianKernel: Gaussian kernel class.

class LaplacianKernel: Laplacian kernel class.

class MMDLoss: Maximum Mean Discrepancy between predictions on two groups of examples.

class MinDiffKernel: MinDiffKernel abstract base class.

class MinDiffLoss: MinDiffLoss abstract base class.