On the efficiency of adaptive MCMC algorithms

被引:18
作者
Andrieu, Christophe [1 ]
Atchade, Yves F.
机构
[1] Univ Bristol, Dept Math, Bristol BS8 1TW, Avon, England
[2] Univ Michigan, Dept Stat, Ann Arbor, MI 48109 USA
来源
ELECTRONIC COMMUNICATIONS IN PROBABILITY | 2007年 / 12卷
关键词
D O I
10.1214/ECP.v12-1320
中图分类号
O21 [概率论与数理统计]; C8 [统计学];
学科分类号
020208 ; 070103 ; 0714 ;
摘要
We study a class of adaptive Markov Chain Monte Carlo (MCMC) processes which aim at behaving as an "optimal" target process via a learning procedure. We show, under appropriate conditions, that the adaptive MCMC chain and the "optimal" (nonadaptive) MCMC processs hare many a symptotic properties. The special case of adaptive MCMC algorithms governed by stochastic approximation is considered in details and we apply our results to the adaptive Metropolis algorithm of Haario et al. (2001).
引用
收藏
页码:336 / 349
页数:14
相关论文
共 13 条