API Reference

The complete biquality-learn project is automatically documented for every module.

Biquality Classifiers

baseline.BiqualityBaseline(estimator[, baseline])

A Biquality Baseline.

density_ratio.KKMM(estimator, *[, kernel, ...])

A K-KMM Density Ratio Biquality Classifier.

density_ratio.KPDR(estimator, *[, ...])

A K-Probabilistic Density Ratio Biquality Classifier.

density_ratio.IKMM(estimator, *[, ...])

An Iterative KMM Density Ratio Biquality Classifier.

density_ratio.IPDR(estimator, *[, ...])

An Iterative Probabilistic Density Ratio Biquality Classifier.

irbl.IRBL(base_estimator, final_estimator)

A Reweighted Classifier for Biquality Learning.

irlnl.IRLNL(base_estimator, final_estimator, *)

A Reweighted Classifier for Learning with Noisy Label [R2e1bf2512a9a-1].

unbiased.LossCorrection(estimator, *[, ...])

A Classifier corrected with the method of unbiased estimators [Rfdaa5ac6e596-1].

plugin.PluginCorrection(estimator, *[, ...])

A Noise Corrected Plug-in Classifier.

tradaboost.TrAdaBoostClassifier([estimator, ...])

A TrAdaBoost classifier.

ea.EasyADAPT(estimator)

A Frustratingly Easy approach to Domain Adaptation.

unhinged.LinearUnhinged(*[, alpha])

Linear Unhinged Classification.

unhinged.KernelUnhinged(*[, alpha, kernel, ...])

Kernel Unhinged Classification.

Transition Matrix Estimators

anchor_transition_matrix(y_prob[, quantile, ...])

Compute the anchor transition matrix [R31489883a896-1].

iterative_anchor_transition_matrix(y_prob[, ...])

Compute a transition matrix based on an iterative algorithm using anchor points [Rfecd00e75182-1].

gold_transition_matrix(y_true, y_prob[, labels])

Compute the gold transition matrix [R059bf0a8afad-1].

Biquality Cross-Validation

make_biquality_cv(X, sample_quality[, cv, ...])

Utility function for building a biquality cross-validator.

Corruptions

make_label_noise(y[, noise_matrix, ...])

Corrupt the labels given a noise transition matrix.

make_instance_dependent_label_noise(...[, ...])

Corrupt the labels given a noise transition matrix and a noise probability function.

uncertainty_noise_probability(X, estimator)

Get a probability of a sample to be noisy given an uncertainty function according to [R82c484bd5ea1-1].

noisy_leaves_probability(X, y, *[, ...])

Noisify some leaves of a decision tree learn on the input dataset.

make_weak_labels(X, y[, estimator, ...])

Generate weak labels for a given dataset.

make_feature_dependent_label_noise(X, y, *)

Corrupt the labels using a noise distribution model by a random linear projection from the features to the labels [Rce44028b087b-1].

make_imbalance(y, *arrays[, majority_ratio, ...])

Create class imbalance in a multi class scenario according to [R748c9ae1839e-1].

make_cluster_imbalance(X, y, *arrays[, ...])

Create per-class cluster imbalance in a multi class scenario according to [R74c8fe2713e2-1].

make_sampling_biais(X, *arrays[, a, b, ...])

Synthetic covariate shift by creating a sampling biais using the first axis of a PCA learned of the input features [R18c441230adf-1].

Noise Matrices

uniform_noise_matrix(n_classes, noise_ratio)

Uniform noise matrix

flip_noise_matrix(n_classes, noise_ratio[, ...])

Flip noise matrix.

background_noise_matrix(n_classes, noise_ratio)

Background noise matrix.