Decomposition and manifold learning

PCA, SVD, matrix factorizations and embeddings for visualization.

AlternatingLeastSquares	Implicit-feedback matrix factorization by alternating least squares.
CCA	Canonical correlation analysis.
ClassicalMDS	Classical (Torgerson) multidimensional scaling.
DictionaryLearning	Learn a dictionary in which the data has a sparse representation.
EllipticEnvelope	Outlier detection for Gaussian-distributed data.
FactorAnalysis	Factor analysis: a linear Gaussian latent variable model.
FastICA	Independent component analysis with the FastICA algorithm.
GaussianRandomProjection	Reduce dimension with a Gaussian random matrix.
IncrementalPCA	Incremental principal component analysis, fitted batch by batch.
Isomap	Isomap embedding: kernel PCA on geodesic distances.
johnson_lindenstrauss_min_dim	Minimum dimension for a random projection that keeps distances within eps.
LatentDirichletAllocation	Latent Dirichlet allocation topic model, fitted by variational EM.
LocallyLinearEmbedding	Locally linear embedding.
lstsq	Least-squares solution of a @ x = b.
lu_factor	LU factorization of a square matrix with partial pivoting.
lu_solve	Solve a linear system from an LU factorization.
manifold	Manifold learning: spectral embedding (Laplacian eigenmaps) and UMAP.
MDS	Multidimensional scaling by SMACOF.
MinCovDet	Minimum covariance determinant: a robust covariance estimator.
MiniBatchDictionaryLearning	Learn a sparse-coding dictionary from mini-batches.
MiniBatchSparsePCA	Sparse principal component analysis fitted on mini-batches.
NMF	Non-negative matrix factorization X ~ W @ H.
PCA	Principal component analysis.
PLSCanonical	Partial least squares in canonical form.
PLSRegression	Partial least squares regression.
randomized_svd	Truncated SVD by randomized range finding.
solve	Solve the square linear system a @ x = b.
sparse_encode	Sparse-code each row of X against a fixed dictionary.
SparseCoder	Sparse coding against a fixed, precomputed dictionary.
SparsePCA	Sparse principal component analysis.
SparseRandomProjection	Reduce dimension with a sparse random matrix.
TruncatedSVD	Dimensionality reduction by truncated singular value decomposition.
UMAP	Uniform Manifold Approximation and Projection (UMAP) embedding.
umap	UMAP: Uniform Manifold Approximation and Projection.