Maxima Units Search (MUS) algorithm: methodology and applications

TL;DR

The paper introduces the Maxima Units Search (MUS) algorithm, a method for extracting identity submatrices and pivotal units from large sparse matrices, with applications including solving label switching in Bayesian models.

Contribution

It presents a novel algorithm for identifying identity submatrices and pivotal units, expanding its application beyond previous uses in Bayesian mixture models.

Findings

Successfully applied to label switching problem in Bayesian models

Effective in extracting identity submatrices from large sparse matrices

Potential for diverse applications in matrix analysis

Abstract

An algorithm for extracting identity submatrices of small rank and pivotal units from large and sparse matrices is proposed. The procedure has already been satisfactorily applied for solving the label switching problem in Bayesian mixture models. Here we introduce it on its own and explore possible applications in different contexts.