metric_learn package

Module Contents

Base Classes

metric_learn.Constraints(partial_labels) Class to build constraints from labeled data.
metric_learn.base_metric.BaseMetricLearner([…]) Base class for all metric-learners.
metric_learn.base_metric._PairsClassifierMixin([…]) Base class for pairs learners.
metric_learn.base_metric._TripletsClassifierMixin([…]) Base class for triplets learners.
metric_learn.base_metric._QuadrupletsClassifierMixin([…]) Base class for quadruplets learners.

Supervised Learning Algorithms

metric_learn.LFDA([n_components, k, …]) Local Fisher Discriminant Analysis for Supervised Dimensionality Reduction
metric_learn.LMNN([init, k, min_iter, …]) Large Margin Nearest Neighbor (LMNN)
metric_learn.MLKR([n_components, init, tol, …]) Metric Learning for Kernel Regression (MLKR)
metric_learn.NCA([init, n_components, …]) Neighborhood Components Analysis (NCA)
metric_learn.RCA([n_components, preprocessor]) Relevant Components Analysis (RCA)
metric_learn.ITML_Supervised([gamma, …]) Supervised version of Information Theoretic Metric Learning (ITML)
metric_learn.LSML_Supervised([tol, …]) Supervised version of Least Squared-residual Metric Learning (LSML)
metric_learn.MMC_Supervised([max_iter, …]) Supervised version of Mahalanobis Metric for Clustering (MMC)
metric_learn.SDML_Supervised([…]) Supervised version of Sparse Distance Metric Learning (SDML)
metric_learn.RCA_Supervised([n_components, …]) Supervised version of Relevant Components Analysis (RCA)
metric_learn.SCML_Supervised([k_genuine, …]) Supervised version of Sparse Compositional Metric Learning (SCML)

Weakly Supervised Learning Algorithms

metric_learn.ITML([gamma, max_iter, …]) Information Theoretic Metric Learning (ITML)
metric_learn.LSML([tol, max_iter, prior, …]) Least Squared-residual Metric Learning (LSML)
metric_learn.MMC([max_iter, max_proj, …]) Mahalanobis Metric for Clustering (MMC)
metric_learn.SDML([balance_param, …]) Sparse Distance Metric Learning (SDML)
metric_learn.SCML([beta, basis, n_basis, …]) Sparse Compositional Metric Learning (SCML)

Unsupervised Learning Algorithms

metric_learn.Covariance([preprocessor]) Covariance metric (baseline method)