EM_lasso: The EM algorithm for the Spike and Slab model inference

In garyganuiuc/SSLasso: Sparse Precision Matrix Estimation with Bayesian Regularization

Usage

EM_lasso(S, n, p_n, v0, v1, maxiter, p, tau)

Arguments

S	Sample Covariance Matrix
n	Number of observations
p_n	Number of variabes/ncol in the precision matrix
v0	The spike tuning parameter
v1	The slab tuning parameter
maxiter	Max iteration ties
p	eta in the paper: the prior probability of non sparsity
tau	The penalty on the diagnoal term

Value

P	Marginal posterior probabilty matrix of each entry being sparse
Theta	An MAP estimate of the precision matrix

garyganuiuc/SSLasso documentation built on May 16, 2019, 5:43 p.m.