Conditions for rapid mixing of parallel and simulated tempering on multimodal distributions

DOI10.1214/08-AAP555zbMATH Open1171.65008OpenAlexW2001464954MaRDI QIDQ1024896FDOQ1024896

Authors D. B. Woodard, Scott C. Schmidler, Mark Huber

Publication date 17 June 2009

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

Full work available at URL https://arxiv.org/abs/0906.2341

zbMATH Keywords

Markov chain Monte Carlo methodspectral gaptemperingMetropolis-coupled algorithmrapidly mixing Markov chains

Mathematics Subject Classification ID

Computational methods in Markov chains (60J22) Monte Carlo methods (65C05) Numerical analysis or methods applied to Markov chains (65C40)

Abstract: We give conditions under which a Markov chain constructed via parallel or simulated tempering is guaranteed to be rapidly mixing, which are applicable to a wide range of multimodal distributions arising in Bayesian statistical inference and statistical mechanics. We provide lower bounds on the spectral gaps of parallel and simulated tempering. These bounds imply a single set of sufficient conditions for rapid mixing of both techniques. A direct consequence of our results is rapid mixing of parallel and simulated tempering for several normal mixture models, and for the mean-field Ising model.

Recommendations

• Sufficient conditions for torpid mixing of parallel and simulated tempering
• Simulated tempering and swapping on mean-field models
• Torpid mixing of simulated tempering on the Potts model
• Accelerating parallel tempering: quantile tempering algorithm (QuanTA)
• Small-world MCMC and convergence to multi-modal distributions: from slow mixing to fast mixing

Cites work

• scientific article; zbMATH DE number 1350773 (Why is no real title available?)
• scientific article; zbMATH DE number 201032 (Why is no real title available?)
• scientific article; zbMATH DE number 840151 (Why is no real title available?)
• Annealing Markov Chain Monte Carlo with Applications to Ancestral Inference
• Bounds on the L 2 Spectrum for Markov Chains and Markov Processes: A Generalization of Cheeger's Inequality
• Comparison theorems for reversible Markov chains
• General state space Markov chains and MCMC algorithms
• Geometric L² and ^{L1 convergence are equivalent for reversible Markov chains}
• Geometric bounds for eigenvalues of Markov chains
• Geometric ergodicity and hybrid Markov chains
• Importance sampling for families of distributions
• Improved Bounds for Mixing Rates of Markov Chains and Multicommodity Flow
• Logarithmic Sobolev inequalities for finite Markov chains
• Markov chain decomposition for convergence rate analysis
• Markov chains for exploring posterior distributions. (With discussion)
• Monte Carlo simulation in statistical physics. An introduction.
• On swapping and simulated tempering algorithms.
• On the swapping algorithm
• Rates of convergence for Gibbs sampling for variance component models
• Sufficient conditions for torpid mixing of parallel and simulated tempering
• Torpid mixing of simulated tempering on the Potts model

Cited in

(37)

• Explaining transmission rate variations and forecasting epidemic spread in multiple regions with a semiparametric mixed effects SIR model
• Optimal scaling of random-walk Metropolis algorithms on general target distributions
• Convergence rate of Markov chain methods for genomic motif discovery
• Generalized parallel tempering on Bayesian inverse problems
• Reflections on Bayesian inference and Markov chain Monte Carlo
• Accelerating MCMC algorithms
• A bound for the convergence rate of parallel tempering for sampling restricted Boltzmann machines
• Spectral gap of replica exchange Langevin diffusion on mixture distributions
• Random Construction of Interpolating Sets for High-Dimensional Integration
• Accelerating parallel tempering: quantile tempering algorithm (QuanTA)
• Parallel hierarchical sampling: a general-purpose interacting Markov chains Monte Carlo algorithm
• Fast mixing of Metropolis-Hastings with unimodal targets
• scientific article; zbMATH DE number 7415116 (Why is no real title available?)
• Small-world MCMC and convergence to multi-modal distributions: from slow mixing to fast mixing
• Simple conditions for metastability of continuous Markov chains
• On swapping and simulated tempering algorithms.
• Complexity results for MCMC derived from quantitative bounds
• On the convergence rates of some adaptive Markov chain Monte Carlo algorithms
• Finite-sample complexity of sequential Monte Carlo estimators
• Error bounds for sequential Monte Carlo samplers for multimodal distributions
• The use of a single pseudo-sample in approximate Bayesian computation
• Distance between configurations in Markov chain Monte Carlo simulations
• Sufficient conditions for torpid mixing of parallel and simulated tempering
• On the swapping algorithm
• Tuning tempered transitions
• Mixing and hitting times for Gibbs samplers and other non-Feller processes
• Mixing times for the swapping algorithm on the Blume-Emery-Griffiths model
• Complexity bounds for Markov chain Monte Carlo algorithms via diffusion limits
• Skew Brownian motion and complexity of the ALPS algorithm
• Weight-preserving simulated tempering
• Approximate spectral gaps for Markov chain mixing times in high dimensions
• Finite sample complexity of sequential Monte Carlo estimators on multimodal target distributions
• Simulated tempering and swapping on mean-field models
• Efficient Multimodal Sampling via Tempered Distribution Flow
• Towards optimal scaling of Metropolis-coupled Markov chain Monte Carlo
• Analysis and optimization of certain parallel Monte Carlo methods in the low temperature limit
• State-dependent swap strategies and automatic reduction of number of temperatures in adaptive parallel tempering algorithm

This page was built for publication: Conditions for rapid mixing of parallel and simulated tempering on multimodal distributions

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1024896)