Efficient MCMC for Gibbs Random Fields using pre-computation

Files in This Item:
File Description SizeFormat 
insight_publication.pdf694.99 kBAdobe PDFDownload
Title: Efficient MCMC for Gibbs Random Fields using pre-computation
Authors: Boland, AidanFriel, NialMarie, Florian
Permanent link: http://hdl.handle.net/10197/10866
Date: 31-May-2018
Online since: 2019-07-09T11:19:29Z
Abstract: Bayesian inference of Gibbs random fields (GRFs) is often referred to as a doubly intractable problem, since the likelihood function is intractable. The exploration of the posterior distribution of such models is typically carried out with a sophisticated Markov chain Monte Carlo (MCMC) method, the exchange algorithm (Murray et al., 2006), which requires simulations from the likelihood function at each iteration. The purpose of this paper is to consider an approach to dramatically reduce this computational overhead. To this end we introduce a novel class of algorithms which use realizations of the GRF model, simulated offline, at locations specified by a grid that spans the parameter space. This strategy speeds up dramatically the posterior inference, as illustrated on several examples. However, using the pre-computed graphs introduces a noise in the MCMC algorithm, which is no longer exact. We study the theoretical behaviour of the resulting approximate MCMC algorithm and derive convergence bounds using a recent theoretical development on approximate MCMC methods.
Funding Details: Science Foundation Ireland
Type of material: Journal Article
Publisher: The Institute of Mathematical Statistics and the Bernoulli Society
Journal:  Electronic Journal of Statistics
Volume: 12
Issue: 2
Start page: 4138
End page: 4179
Copyright (published version): 2018 the Authors
Keywords: Machine Learning & StatisticsBayesianGibbs random fields (GRFs)Markov chain Monte Carlo (MCMC) methodAlgorithmsGRF modelMCMC algorithm
DOI: 10.1214/18-EJS1504
Language: en
Status of Item: Peer reviewed
Appears in Collections:Insight Research Collection

Show full item record

Google ScholarTM



This item is available under the Attribution-NonCommercial-NoDerivs 3.0 Ireland. No item may be reproduced for commercial purposes. For other possible restrictions on use please refer to the publisher's URL where this is made available, or to notes contained in the item itself. Other terms may apply.