README.md

In Bo-Ning/VBsparsePCA: The Variational Bayesian Method for Sparse PCA

VBsparsePCA: The Variational Bayesian Method for Sparse PCA

This package contains functions for a variational Bayesian approach for sparse PCA. The algorithm is the PX-CAVI algorithm proposed by Ning (2021) (arXiv:2102.00305) if assuming the jointly row-sparsity assumption for the loadings matrix and the batch PX-CAVI algorithm if otherwise. The outputs of the main function, VBsparsePCA, include the mean and covariance of the loadings matrix, the score functions, the variable selection results, and the estimated variance of the random noise.

See VBsparsePCA.pdf for details.

Bo-Ning/VBsparsePCA documentation built on Feb. 14, 2021, 6:10 p.m.