Using Parallel Computation to Improve Independent Metropolis–Hastings Based Estimation

In this article, we consider the implications of the fact that parallel raw-power can be exploited by a generic Metropolis–Hastings algorithm if the proposed values are independent from the current value of the Markov chain. In particular, we present improvements to the independent Metropolis–Hastings algorithm that significantly decrease the variance of any estimator derived from the MCMC output, at a null computing cost since those improvements are based on a fixed number of target density evaluations that can be produced in parallel. The techniques developed in this article do not jeopardize the Markovian convergence properties of the algorithm, since they are based on the Rao–Blackwell principles of Gelfand and Smith (1990), already exploited in the work of Casella and Robert (1996), Atchadé and Perron (2005), and Douc and Robert (2011). We illustrate those improvements both on a toy normal example and on a classical probit regression model, but stress the fact that they are applicable in any case where the independent Metropolis–Hastings is applicable. The code used in this article is available as supplementary material.

[1]  François Perron,et al.  Improving on the Independent Metropolis-Hastings algorithm , 2005 .

[2]  Mats Gyllenberg,et al.  Bayesian model learning based on a parallel MCMC strategy , 2006, Stat. Comput..

[3]  C. Robert,et al.  Importance sampling methods for Bayesian discrimination between embedded models , 2009, 0910.2325.

[4]  Sally Rosenthal,et al.  Parallel computing and Monte Carlo algorithms , 1999 .

[5]  G. Casella,et al.  Rao-Blackwellisation of sampling schemes , 1996 .

[6]  Galin L. Jones,et al.  On the applicability of regenerative simulation in Markov chain Monte Carlo , 2002 .

[7]  F Perron Beyond accept-reject sampling , 1999 .

[8]  N. Chopin A sequential particle filter method for static models , 2002 .

[9]  Bin Yu,et al.  Regeneration in Markov chain samplers , 1995 .

[10]  S. Chib,et al.  Bayesian analysis of binary and polychotomous response data , 1993 .

[11]  Pierre L'Ecuyer,et al.  An Object-Oriented Random-Number Package with Many Long Streams and Substreams , 2002, Oper. Res..

[12]  Cliburn Chan,et al.  Understanding GPU Programming for Statistical Computation: Studies in Massively Parallel Massive Mixtures , 2010, Journal of computational and graphical statistics : a joint publication of American Statistical Association, Institute of Mathematical Statistics, Interface Foundation of North America.

[13]  C. Robert Convergence Control Methods for Markov Chain Monte Carlo Algorithms , 1995 .

[14]  W. Gilks,et al.  Adaptive Rejection Metropolis Sampling Within Gibbs Sampling , 1995 .

[15]  Jeffrey S. Rosenthal,et al.  Adaptive Gibbs samplers , 2010 .

[16]  Christian P. Robert,et al.  A vanilla Rao--Blackwellization of Metropolis--Hastings algorithms , 2009, 0904.2144.

[17]  Chao Yang,et al.  Learn From Thy Neighbor: Parallel-Chain and Regional Adaptive MCMC , 2009 .

[18]  Radu V. Craiu,et al.  Multiprocess parallel antithetic coupling for backward and forward Markov Chain Monte Carlo , 2005, math/0505631.

[19]  Roman Holenstein,et al.  Particle Markov chain Monte Carlo , 2009 .

[20]  Arnaud Doucet,et al.  On the Utility of Graphics Cards to Perform Massively Parallel Simulation of Advanced Monte Carlo Methods , 2009, Journal of computational and graphical statistics : a joint publication of American Statistical Association, Institute of Mathematical Statistics, Interface Foundation of North America.

[21]  Hoon Kim,et al.  Monte Carlo Statistical Methods , 2000, Technometrics.

[22]  Adrian F. M. Smith,et al.  Sampling-Based Approaches to Calculating Marginal Densities , 1990 .

[23]  P. Moral,et al.  Sequential Monte Carlo samplers , 2002, cond-mat/0212648.

[24]  Gersende Fort,et al.  Estimation of cosmological parameters using adaptive importance sampling , 2009, 0903.0837.