Optimal transport couplings of Gibbs samplers on partitions for unbiased estimation

04/09/2021
by   Brian L Trippe, et al.
0

Computational couplings of Markov chains provide a practical route to unbiased Monte Carlo estimation that can utilize parallel computation. However, these approaches depend crucially on chains meeting after a small number of transitions. For models that assign data into groups, e.g. mixture models, the obvious approaches to couple Gibbs samplers fail to meet quickly. This failure owes to the so-called "label-switching" problem; semantically equivalent relabelings of the groups contribute well-separated posterior modes that impede fast mixing and cause large meeting times. We here demonstrate how to avoid label switching by considering chains as exploring the space of partitions rather than labelings. Using a metric on this space, we employ an optimal transport coupling of the Gibbs conditionals. This coupling outperforms alternative couplings that rely on labelings and, on a real dataset, provides estimates more precise than usual ergodic averages in the limited time regime. Code is available at github.com/tinnguyen96/coupling-Gibbs-partition.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
02/23/2022

Many processors, little time: MCMC for partitions via optimal transport couplings

Markov chain Monte Carlo (MCMC) methods are often used in clustering sin...
research
11/05/2019

Alleviating Label Switching with Optimal Transport

Label switching is a phenomenon arising in mixture model posterior infer...
research
06/14/2020

Optimal Transport for Stationary Markov Chains via Policy Iteration

We study an extension of optimal transport techniques to stationary Mark...
research
02/02/2021

Couplings of the Random-Walk Metropolis algorithm

Couplings play a central role in contemporary Markov chain Monte Carlo m...
research
09/01/2017

Unbiased Hamiltonian Monte Carlo with couplings

We propose a coupling approach to parallelize Hamiltonian Monte Carlo es...
research
11/08/2022

Scalable multi-class sampling via filtered sliced optimal transport

We propose a multi-class point optimization formulation based on continu...
research
10/06/2009

Distance Dependent Chinese Restaurant Processes

We develop the distance dependent Chinese restaurant process (CRP), a fl...

Please sign up or login with your details

Forgot password? Click here to reset