A Gibbs Sampling Alternative to Reversible Jump MCMC

TL;DR

This paper introduces a straightforward Gibbs sampling method as an alternative to reversible jump MCMC for Bayesian model selection and trans-dimensional sampling.

Contribution

It proposes a simple and elegant Gibbs sampling approach that replaces the more complex reversible jump MCMC technique.

Findings

Offers a practical alternative to reversible jump MCMC

Simplifies implementation of trans-dimensional sampling

Potentially improves computational efficiency

Abstract

This note presents a simple and elegant sampler which could be used as an alternative to the reversible jump MCMC methodology.