Composite likelihood inference in a discrete latent variable model for two-way "clustering-by-segmentation" problems

TL;DR

This paper introduces a composite likelihood approach for a complex discrete latent variable model that simultaneously clusters along one data dimension and segments along another, applicable to two-way data arrays.

Contribution

It develops a novel composite likelihood methodology for a complex hidden Markov model in two-way clustering-by-segmentation problems, enabling feasible inference.

Findings

Effective in simulation studies

Successfully applied to genomic data

Provides a practical alternative to full likelihood estimation

Abstract

We consider a discrete latent variable model for two-way data arrays, which allows one to simultaneously produce clusters along one of the data dimensions (e.g. exchangeable observational units or features) and contiguous groups, or segments, along the other (e.g. consecutively ordered times or locations). The model relies on a hidden Markov structure but, given its complexity, cannot be estimated by full maximum likelihood. We therefore introduce composite likelihood methodology based on considering different subsets of the data. The proposed approach is illustrated by simulation, and with an application to genomic data.