Direction-sweep Markov chains

DOI10.1088/1751-8121/AC508AzbMATH Open1505.82066arXiv2007.15615OpenAlexW4210270034MaRDI QIDQ5049692FDOQ5049692

Authors: Liang Qin, Philipp Höllmer, Werner Krauth 

Publication date: 11 November 2022

Published in: Journal of Physics A: Mathematical and Theoretical (Search for Journal in Brave)

Abstract: We discuss a non-reversible, lifted Markov-chain Monte Carlo (MCMC) algorithm for particle systems in which the direction of proposed displacements is changed deterministically. This algorithm sweeps through directions analogously to the popular MCMC sweep methods for particle or spin indices. Direction-sweep MCMC can be applied to a wide range of original reversible or non-reversible Markov chains, such as the Metropolis algorithm or the event-chain Monte Carlo algorithm. For a single two-dimensional dipole, we consider direction-sweep MCMC in the limit where restricted equilibrium is reached among the accessible configurations before changing the direction. We show rigorously that direction-sweep MCMC leaves the stationary probability distribution unchanged, and that it profoundly modifies the Markov-chain trajectory. Long excursions, with persistent rotation in one direction, alternate with long sequences of rapid zigzags resulting in persistent rotation in the opposite direction in the limit of small direction increments. The mapping to a Langevin equation then yields the exact scaling of excursions while the zigzags are described through a non-linear differential equation that is solved exactly. We show that the direction-sweep algorithm can have shorter mixing times than the algorithms with random updates of directions. We point out possible applications of direction-sweep MCMC in polymer physics and in molecular simulation.

Full work available at URL: https://arxiv.org/abs/2007.15615

Recommendations

• A note on Markov chain Monte Carlo sweep strategies
• Markov chain Monte Carlo and irreversibility
• PDMP characterisation of event-chain Monte Carlo algorithms for particle systems
• Irreversible Monte Carlo algorithms for efficient sampling
• Markov chain Monte Carlo algorithms allowing parallel processing. II

zbMATH Keywords

stochastic processesMonte Carlo methodsLangevin equationnon-reversible Markov chains

Mathematics Subject Classification ID

Computational methods in Markov chains (60J22) Monte Carlo methods applied to problems in statistical mechanics (82M31)

Cites Work

• Analysis of a nonreversible Markov chain sampler.
• Equation of state calculations by fast computing machines
• Irreversible Monte Carlo algorithms for efficient sampling
• Probability Theory
• The Fokker-Planck equation. Methods of solution and applications.
• Statistical mechanics: Algorithms and computations. With CD-ROM.
• Title not available (Why is that?)
• The mathematics of mixing things up
• A piecewise deterministic scaling limit of lifted Metropolis-Hastings in the Curie-Weiss model
• Non-reversible Monte Carlo simulations of spin models
• Piecewise deterministic Markov processes for scalable Monte Carlo on restricted domains
• Forward Event-Chain Monte Carlo: Fast Sampling by Randomness Control in Irreversible Markov Chains

Cited In (1)

• PDMP characterisation of event-chain Monte Carlo algorithms for particle systems

This page was built for publication: Direction-sweep Markov chains

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5049692)