arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1602.02889stat.MEmath.STstat.TH
收藏

可逆方案下马尔可夫链蒙特卡罗的遍历性

Ergodicity of Markov chain Monte Carlo with reversible proposal

Kengo Kamatani

摘要:

我们描述了一些Metropolis-Hastings(MH)算法的遍历性质的重尾目标分布。分析通常属于亚几何遍历框架,但我们证明了混合预处理Crank-Nicolson(MpCN)算法具有几何遍历性,即使是重尾目标分布。这个有用的属性来自于MpCN算法在适当的变换下变成随机行走Metropolis算法的事实。

英文摘要:

We describe ergodic properties of some Metropolis-Hastings (MH) algorithms for heavy-tailed target distributions. The analysis usually falls into sub-geometric ergodicity framework but we prove that the mixed preconditioned Crank-Nicolson (MpCN) algorithm has geometric ergodicity even for heavy-tailed target distributions. This useful property comes from the fact that the MpCN algorithm becomes a random-walk Metropolis algorithm under suitable transformation.

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。

↑