arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2211.03487cs.DScs.CCcs.DMmath.PR

迈向马尔可夫链蒙特卡洛的去随机化

Towards derandomising Markov chain Monte Carlo

Weiming Feng, Heng Guo, Chunyang Wang, Jiaheng Wang, Yitong Yin

AI总结:

本文提出了一种基于向后耦合的去随机化框架,通过将对数级随机选择实现对马尔可夫链蒙特卡洛算法的去随机化,并成功应用于超图独立集与超图染色的确定性近似计数。

摘要:

我们提出了一个新框架来去随机化某些马尔可夫链蒙特卡洛(MCMC)算法。如同在MCMC中,我们首先将计数问题归约为从一系列边缘分布中采样。对于后一项任务,我们引入了一种称为“向后耦合”的方法,该方法能在对数时间内,从平稳马尔可夫链状态中求出一个或常数个变量的值。由于最多只有对数级的随机选择,这带来了非常简单的去随机化。我们给出了该框架的两个应用,即在局部引理类型条件下,针对超图独立集和超图染色的确定性近似计数算法,在忽略低阶因子的情况下,其效率匹配了目前最先进的随机化对应算法。

英文摘要:

We present a new framework to derandomise certain Markov chain Monte Carlo (MCMC) algorithms. As in MCMC, we first reduce counting problems to sampling from a sequence of marginal distributions. For the latter task, we introduce a method called coupling towards the past that can, in logarithmic time, evaluate one or a constant number of variables from a stationary Markov chain state. Since there are at most logarithmic random choices, this leads to very simple derandomisation. We provide two applications of this framework, namely efficient deterministic approximate counting algorithms for hypergraph independent sets and hypergraph colourings, under local lemma type conditions matching, up to lower order factors, their state-of-the-art randomised counterparts.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑