MRCpy Package Contents

Numpy-based Classifiers

MRCpy.MRC([loss, s, deterministic, ...])	Minimax Risk Classifier
MRCpy.CMRC([loss, s, deterministic, ...])	Constrained Minimax Risk Classifier
MRCpy.AMRC(n_classes[, loss, deterministic, ...])	Adaptive Minimax Risk Classifier
MRCpy.DWGCS([loss, deterministic, ...])	Double-Weighting for General Covariate Shift
MRCpy.LMRC([s, loss, n_classes, alpha, eps, ...])	Cost Sensitive Minimax Risk Classifier.
MRCpy.LCMRC([s, loss, n_classes, alpha, ...])	Marginally Constrained Cost Sensitive Minimax Risk Classifier.

PyTorch-based Classifiers

MRCpy.pytorch.mgce.classifier.mgce_clf([...])	Minimax Generalized Cross-Entropy (MGCE).
MRCpy.pytorch.mgce.loss.mgce_loss(num_classes)	Margin Loss for Minimax Generalized Cross-Entropy (MGCE) Classification.

Feature Mappings

MRCpy.phi.RandomFourierPhi(n_classes[, ...])	Fourier features
MRCpy.phi.RandomReLUPhi(n_classes[, ...])	ReLU features
MRCpy.phi.ThresholdPhi(n_classes[, ...])	Threshold features

Extended Functionalities

MRCpy.BaseMRC([loss, s, deterministic, ...])	Base class for different minimax risk classifiers
MRCpy.phi.BasePhi(n_classes[, ...])	Base class for feature mappings