VBsparsePCA: The Variational Bayesian Method for Sparse PCA

Contains functions for a variational Bayesian method for sparse PCA proposed by Ning (2020) <arXiv:2102.00305>. There are two algorithms: the PX-CAVI algorithm (if assuming the loadings matrix is jointly row-sparse) and the batch PX-CAVI algorithm (if without this assumption). 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.

Getting started

Package details

AuthorBo (Yu-Chien) Ning
MaintainerBo (Yu-Chien) Ning <bo.ning@upmc.fr>
Package repositoryView on CRAN
Installation Install the latest version of this package by entering the following in R:

Try the VBsparsePCA package in your browser

Any scripts or data that you put into this service are public.

VBsparsePCA documentation built on Feb. 12, 2021, 5:06 p.m.