Potential-Decomposition Strategy in Markov Chain Monte Carlo Sampling Algorithms

• 来源期刊: 理论物理通讯(英文版)
• 学科: 数学
• 关键词: 马尔可夫链 蒙特卡罗 算法 抽样 分解 Ising模型 综合布线 试验样品
• 年，卷（期）: 2010,（11）
• 所属期刊栏目: Geophysics, Astronomy, and Astrophysics
• 页码范围: 854-856
• 页数: 3页
• 分类号: O211.62 TL329.2
• 语种: 英文

摘要

We introduce the potential-decomposition strategy (PDS), which can be used in Markov chain Monte Carlosampling algorithms.PDS can be designed to make particles move in a modified potential that favors diffusion in phasespace, then, by rejecting some trial samples, the target distributions can be sampled in an unbiased manner.Furthermore,if the accepted trial samples are insufficient, they can be recycled as initial states to form more unbiased samples.Thisstrategy can greatly improve efficiency when the original potential has multiple metastable states separated by largebarriers.We apply PDS to the 2d Ising model and a double-well potential model with a large barrier, demonstrating inthese two representative examples that convergence is accelerated by orders of magnitude.