2014

Non-reversible Metropolis-Hastings

Bierkens, Joris

Understand

The classical Metropolis-Hastings (MH) algorithm can be extended to generate non-reversible Markov chains.

This is achieved by means of a modification of the acceptance probability, using the notion of vorticity matrix.
The resulting Markov chain is non-reversible.
Results from the literature on asymptotic variance, large deviations theory and mixing time are mentioned, and in the case of a large deviations result, adapted, to explain how non-reversible Markov chains have favorable properties in these respects.

Equation of State Calculations by Fast Computing Machines
Nicholas Metropolis, Arianna W. Rosenbluth, Marshall N. Rosenbluth, Augusta H. Teller, and Edward Teller · 1953
Earlier work this paper cites.
Monte Carlo sampling methods using Markov chains and their applications
WK Hastings · 1970
Earlier work this paper cites.
Functional Analysis
Original
Kosaku Yosida · 1980
Earlier work this paper cites.
The theory of matrices
Peter Lancaster and Miron Tismenetsky · 1985
Earlier work this paper cites.
Real and complex analysis
Walter Rudin · 1987
Earlier work this paper cites.
Matrix analysis
Roger A Horn and Charles R Johnson · 1990
Earlier work this paper cites.
A generalized guided Monte Carlo algorithm
Alan M. Horowitz · 1991
Earlier work this paper cites.
Accelerating Gaussian diffusions
CR Hwang, SY Hwang-Ma, and SJ Sheu · 1993
Earlier work this paper cites.
Exponential Convergence of Langevin Distributions and Their Discrete Approximations
Gareth O Roberts and Richard L Tweedie · 1996
Earlier work this paper cites.
What do we know about the Metropolis algorithm?
P Diaconis and L Saloff-Coste · 1998
Earlier work this paper cites.

Optimal scaling of discrete approximations to Langevin diffusions
Gareth O. Roberts and Jeffrey S. Rosenthal · 1998
Cited alongside, same era.
Lifting Markov chains to speed up mixing
Fang Chen, László Lovász, and Igor Pak · 1999
Cited alongside, same era.
Exact rates of convergence for some simple non-reversible Markov chains
EL Wilmer · 1999
Cited alongside, same era.
Large deviations
Frank den Hollander · 2000
Cited alongside, same era.
Analysis of a nonreversible Markov chain sampler
Persi Diaconis, Susan Holmes, and RM Neal · 2000
Cited alongside, same era.
On non-reversible Markov chains
Charles J Geyer and Antonietta Mira · 2000
Cited alongside, same era.
Improving Asymptotic Variance of MCMC Estimators: Non-reversible Chains are Better
Radford M. Neal · 2004
Cited alongside, same era.
General state space Markov chains and MCMC algorithms
Gareth O Roberts and Jeffrey S Rosenthal · 2004
Cited alongside, same era.
The Markov chain Monte Carlo revolution
Persi Diaconis · 2008
Cited alongside, same era.

Markov chains and mixing times
DA Levin, Yuval Peres, and EL Wilmer · 2009
Later among the works it cites.
Improving the Asymptotic Performance of Markov Chain Monte-Carlo by Inserting Vortices
Yi Sun, Faustino Gomez, and Juergen Schmidhuber · 2010
Later among the works it cites.
Markov Chain Monte Carlo Method without Detailed Balance
Hidemaro Suwa and Synge Todo · 2010
Later among the works it cites.
MCMC using Hamiltonian dynamics
R Neal · 2011
Later among the works it cites.
Irreversible Monte Carlo algorithms for efficient sampling
Konstantin S. Turitsyn, Michael Chertkov, and Marija Vucelja · 2011
Later among the works it cites.
Accelerating reversible Markov chains
Ting-Li Chen and Chii-Ruey Hwang · 2013
Later among the works it cites.
Optimal Non-reversible Linear Drift for the Convergence to Equilibrium of a Diffusion
T. Lelièvre, F. Nier, and G. a. Pavliotis · 2013
Later among the works it cites.
Irreversible Langevin samplers and variance reduction: a large deviation approach
Luc Rey-Bellet and Kostantinos Spiliopoulos · 2014
Closest in time.
Lifting – A Nonreversible Markov Chain Monte Carlo Algorithm
Marija Vucelja · 2014
Closest in time.

alphaXiv searches the wider corpus for related work and actual follow-ups.

alphaXiv is searching for related work…

Reading the bibliography…