An irreversible Markov-chain Monte Carlo method with skew detailed balance conditions

ELC International Meeting on Inference, Computation, and Spin Glasses

Hukushima, K.^1 ; Sakai, Y.^1

Department of Basic Science, Graduate School of Arts and Sciences, University of Tokyo, 3-8-1 Komaba, Meguro-ku, Tokyo 153-8902, Japan^1

关键词: Detailed balance;    Dynamical exponents;    Ferromagnetic Ising models;    Markov chain Monte Carlo method;    Metropolis-Hastings algorithm;    Order parameter;    Relaxation dynamics;    Three dimensions;   

Others  :  https://iopscience.iop.org/article/10.1088/1742-6596/473/1/012012/pdf
DOI  :  10.1088/1742-6596/473/1/012012

来源: IOP

【 摘 要 】

An irreversible Markov-chain Monte Carlo (MCMC) method based on a skew detailed balance condition is discussed. Some recent theoretical works concerned with the irreversible MCMC method are reviewed and the irreversible Metropolis-Hastings algorithm for the method is described. We apply the method to ferromagnetic Ising models in two and three dimensions. Relaxation dynamics of the order parameter and the dynamical exponent are studied in comparison to those with the conventional reversible MCMC method with the detailed balance condition. We also examine how the efficiency of exchange Monte Carlo method is affected by the combined use of the irreversible MCMC method.