Bayesian sparse convex clustering via global-local shrinkage priors

Abstract Sparse convex clustering is to group observations and conduct variable selection simultaneously in the framework of clustering. Although a weighted $$L_1$$ L 1 norm usually employed for regularization term sparse clustering, its use increases dependence on data reduces estimation accuracy if sample size not sufficient. To tackle these problems, this paper proposes Bayesian method based ideas lasso global-local shrinkage priors. We introduce Gibbs sampling algorithms our using scale mixtures normal distributions. The effectiveness proposed methods shown simulation studies real analysis.