Optimal scaling of random walk Metropolis algorithms with discontinuous target densities

DOI10.1214/11-AAP817zbMATH Open1259.60082OpenAlexW3101802995MaRDI QIDQ691108FDOQ691108

Authors Peter Neal, Wai Kong Yuen, Gareth O. Roberts 

Publication date 29 November 2012

Published in The Annals of Applied Probability (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1210.5090

zbMATH Keywords

Markov chain Monte Carlooptimal scalingdiscontinuous densitiesrandom walk Metropolis algorithm

Mathematics Subject Classification ID

Computational methods in Markov chains (60J22) Monte Carlo methods (65C05) Numerical analysis or methods applied to Markov chains (65C40) Central limit and other weak theorems (60F05) Discrete-time Markov processes on general state spaces (60J05)

Abstract: We consider the optimal scaling problem for high-dimensional random walk Metropolis (RWM) algorithms where the target distribution has a discontinuous probability density function. Almost all previous analysis has focused upon continuous target densities. The main result is a weak convergence result as the dimensionality d of the target densities converges to infinity. In particular, when the proposal variance is scaled by $d^{-2}$, the sequence of stochastic processes formed by the first component of each Markov chain converges to an appropriate Langevin diffusion process. Therefore optimizing the efficiency of the RWM algorithm is equivalent to maximizing the speed of the limiting diffusion. This leads to an asymptotic optimal acceptance rate of $e^{-2}$ (=0.1353) under quite general conditions. The results have major practical implications for the implementation of RWM algorithms by highlighting the detrimental effect of choosing RWM algorithms over Metropolis-within-Gibbs algorithms.

Recommendations

• Weak convergence and optimal scaling of random walk Metropolis algorithms
• Optimal scaling for random walk Metropolis on spherically constrained target densities
• Optimal scaling of Metropolis algorithms: Heading toward general target distributions
• Optimal scaling of random-walk Metropolis algorithms on general target distributions
• Optimal scaling of random walk Metropolis algorithms with non-Gaussian proposals

Cites work

• scientific article; zbMATH DE number 3951715 (Why is no real title available?)
• scientific article; zbMATH DE number 4080508 (Why is no real title available?)
• scientific article; zbMATH DE number 52632 (Why is no real title available?)
• scientific article; zbMATH DE number 3638844 (Why is no real title available?)
• scientific article; zbMATH DE number 2150787 (Why is no real title available?)
• scientific article; zbMATH DE number 3274494 (Why is no real title available?)
• From Metropolis to diffusions: Gibbs states and optimal scaling.
• Optimal acceptance rates for Metropolis algorithms: Moving beyond 0.234
• Optimal scaling for partially updating MCMC algorithms
• Optimal scaling for random walk Metropolis on spherically constrained target densities
• Optimal scaling for various Metropolis-Hastings algorithms.
• Optimal scaling of random walk Metropolis algorithms with non-Gaussian proposals
• Optimal scaling of the random walk Metropolis on elliptically symmetric unimodal targets
• Scaling Limits for the Transient Phase of Local Metropolis–Hastings Algorithms
• Weak convergence and optimal scaling of random walk Metropolis algorithms
• Weak convergence of Metropolis algorithms for non-I.I.D. target distributions

Cited in

(27)

• Optimal scaling of random-walk Metropolis algorithms on general target distributions
• Comparison of hit-and-run, slice sampler and random walk Metropolis
• Optimal scaling for the transient phase of Metropolis Hastings algorithms: the longtime behavior
• Hierarchical models: local proposal variances for RWM-within-Gibbs and MALA-within-Gibbs
• Forward simulation Markov chain Monte Carlo with applications to stochastic epidemic models
• Parallel hierarchical sampling: a general-purpose interacting Markov chains Monte Carlo algorithm
• Optimal scaling of the random walk Metropolis algorithm under L^p mean differentiability
• Sampling constrained probability distributions using spherical augmentation
• The random walk Metropolis: linking theory and practice through a case study
• Optimal scaling of Metropolis algorithms: Heading toward general target distributions
• Sampling constrained continuous probability distributions: a review
• Optimal scaling of random walk Metropolis algorithms with non-Gaussian proposals
• Optimal scaling for the transient phase of the random walk Metropolis algorithm: the mean-field limit
• Hierarchical models and tuning of random walk Metropolis algorithms
• Optimal scaling of random walk Metropolis algorithms using Bayesian large-sample asymptotics
• Asymptotic analysis of the random walk metropolis algorithm on ridged densities
• Optimal scaling for random walk Metropolis on spherically constrained target densities
• Multiplicative random walk Metropolis-Hastings on the real line
• A Dirichlet form approach to MCMC optimal scaling
• Dimension-Independent MCMC Sampling for Inverse Problems with Non-Gaussian Priors
• Automatically tuned general-purpose MCMC via new adaptive diagnostics
• Complexity bounds for Markov chain Monte Carlo algorithms via diffusion limits
• Counterexamples for optimal scaling of Metropolis-Hastings chains with rough target densities
• Diffusion limit for the random walk Metropolis algorithm out of stationarity
• Collapsing of non-centred parameterized MCMC algorithms with applications to epidemic models
• Random walk Metropolis algorithm in high dimension with non-Gaussian target distributions
• Bayesian computation: a summary of the current state, and samples backwards and forwards

This page was built for publication: Optimal scaling of random walk Metropolis algorithms with discontinuous target densities

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q691108)