The Annals of Applied Probability
- Ann. Appl. Probab.
- Volume 28, Number 5 (2018), 2966-3001.
Asymptotic analysis of the random walk Metropolis algorithm on ridged densities
We study the asymptotic behaviour of the Random Walk Metropolis algorithm on “ridged” probability densities where most of the probability mass is distributed along some key directions. Such class of probability measures arise in various applied contexts including for instance Bayesian inverse problems where the posterior measure concentrates on a manifold when the noise variance goes to zero. When the target measure concentrates on a linear manifold, we derive analytically a diffusion limit for the Random Walk Metropolis Markov chain as the scale parameter goes to zero. In contrast to the existing works on scaling limits, our limiting stochastic differential equation does not in general have a constant diffusion coefficient. Our results show that in some cases, the usual practice of adapting the step-size to control the acceptance probability might be sub-optimal as the optimal acceptance probability is zero (in the limit).
Ann. Appl. Probab., Volume 28, Number 5 (2018), 2966-3001.
Received: December 2016
Revised: September 2017
First available in Project Euclid: 28 August 2018
Permanent link to this document
Digital Object Identifier
Mathematical Reviews number (MathSciNet)
Zentralblatt MATH identifier
Primary: 65C05: Monte Carlo methods 65C40: Computational Markov chains
Secondary: 65C30: Stochastic differential and integral equations
Beskos, Alexandros; Roberts, Gareth; Thiery, Alexandre; Pillai, Natesh. Asymptotic analysis of the random walk Metropolis algorithm on ridged densities. Ann. Appl. Probab. 28 (2018), no. 5, 2966--3001. doi:10.1214/18-AAP1380. https://projecteuclid.org/euclid.aoap/1535443239