On the efficiency of pseudo-marginal random walk Metropolis algorithms

Abstract

We examine the behaviour of the pseudo-marginal random walk Metropolis algorithm, where evaluations of the target density for the accept/reject probability are estimated rather than computed precisely. Under relatively general conditions on the target distribution, we obtain limiting formulae for the acceptance rate and for the expected squared jump distance, as the dimension of the target approaches infinity, under the assumption that the noise in the estimate of the log-target is additive and is independent of the position. For targets with independent and identically distributed components, we also obtain a limiting diffusion for the first component. We then consider the overall efficiency of the algorithm, in terms of both speed of mixing and computational time. Assuming the additive noise is Gaussian and is inversely proportional to the number of unbiased estimates that are used, we prove that the algorithm is optimally efficient when the variance of the noise is approximately 3.283 and the acceptance rate is approximately 7.001%. We also find that the optimal scaling is insensitive to the noise and that the optimal variance of the noise is insensitive to the scaling. The theory is illustrated with a simulation study using the particle marginal random walk Metropolis.

Published in at http://dx.doi.org/10.1214/14-AOS1278 the Annals of Statistics (http://www.imstat.org/aos/) by the Institute of Mathematical Statistics (http://www.imstat.org)

References in corpus (8)

• The pseudo-marginal approach for efficient Monte Carlo computations
• Optimal scalings for local Metropolis--Hastings chains on nonproduct targets in high dimensions
• The Random Walk Metropolis: Linking Theory and Practice Through a Case Study
• Optimal scaling of the random walk Metropolis on elliptically symmetric unimodal targets
• Weak convergence of Metropolis algorithms for non-i.i.d. target distributions
• Coupled MCMC with a randomized acceptance probability
• Minimising MCMC variance via diffusion limits, with an application to simulated tempering
• Optimal Scaling of Mala for Nonlinear Regression

Cited by in corpus (44)

• On Particle Methods for Parameter Estimation in State-Space Models
• Speeding Up MCMC by Efficient Data Subsampling
• Three discussions of the paper "sequential quasi-Monte Carlo sampling", by M. Gerber and N. Chopin
• Particle Filters and Data Assimilation
• Speeding Up MCMC by Delayed Acceptance and Data Subsampling
• Metropolized Randomized Maximum Likelihood for sampling from multimodal distributions
• Black-box Bayesian inference for economic agent-based models
• Accelerating pseudo-marginal Metropolis-Hastings by correlating auxiliary variables
• Importance sampling type estimators based on approximate marginal MCMC
• Bayesian inference for stochastic differential equation mixed effects models of a tumor xenography study
• Probabilistic preference learning with the Mallows rank model
• Bayesian Inference for Hybrid Discrete-Continuous Stochastic Kinetic Models
• The impact of temporal sampling resolution on parameter inference for biological transport models
• Coupling stochastic EM and Approximate Bayesian Computation for parameter inference in state-space models
• Bayesian computation: a perspective on the current state, and sampling backwards and forwards
• Augmentation Schemes for Particle MCMC
• Hamiltonian Monte Carlo with Energy Conserving Subsampling
• The chopthin algorithm for resampling
• Comparison of Markov chains via weak Poincaré inequalities with application to pseudo-marginal MCMC
• Importance sampling for partially observed temporal epidemic models
• Particle Metropolis adjusted Langevin algorithms for state space models
• The Neural Moving Average Model for Scalable Variational Inference of State Space Models
• Accelerating delayed-acceptance Markov chain Monte Carlo algorithms
• A survey of Monte Carlo methods for noisy and costly densities with application to reinforcement learning and ABC
• On Coupling Particle Filter Trajectories
• Accelerating sequential Monte Carlo with surrogate likelihoods
• Optimal Scaling of Random-Walk Metropolis Algorithms on General Target Distributions
• Likelihood-free stochastic approximation EM for inference in complex models
• Sequential Monte Carlo Methods in the nimble R Package
• Markov Chain Monte Carlo with the Integrated Nested Laplace Approximation
• Computationally Efficient Bayesian Estimation of High Dimensional Copulas with Discrete and Mixed Margins
• A Common Derivation for Markov Chain Monte Carlo Algorithms with Tractable and Intractable Targets
• Stability of Noisy Metropolis-Hastings
• Subsampling MCMC - An introduction for the survey statistician
• Quasi-Newton particle Metropolis-Hastings
• Change point detection in dynamic Gaussian graphical models: the impact of COVID-19 pandemic on the US stock market
• Unbiased Bayesian Inference for Population Markov Jump Processes via Random Truncations
• Approximate Bayesian Computation for a Class of Time Series Models
• Bayesian Inference of Reproduction Number from Epidemiological and Genetic Data Using Particle MCMC
• Bring the noise: exact inference from noisy simulations in collider physics
• A Pseudo-Marginal Metropolis-Hastings Algorithm for Estimating Generalized Linear Models in the Presence of Missing Data
• Optimal scaling for the pseudo-marginal random walk Metropolis: insensitivity to the noise generating mechanism
• Robust Particle Density Tempering for State Space Models
• The Node-wise Pseudo-marginal Method