A mixture representation of π with applications in Markov chain Monte Carlo and perfect sampling

• 成果类型: Article
• 署名作者: Hobert, JP; Robert, CP
• 署名单位: State University System of Florida; University of Florida; Universite PSL; Universite Paris-Dauphine; Institut Polytechnique de Paris; ENSAE Paris
• 刊物名称: ANNALS OF APPLIED PROBABILITY
• ISSN/ISSBN: 1050-5164
• DOI: 10.1214/105051604000000305
• 发表日期: 2004
• 页码: 1295-1305
• 关键词: CONVERGENCE-RATES

摘要

Let X = {X-n : n = 0, 1, 2, . . .} be an irreducible, positive recurrent Markov chain with invariant probability measure pi. We show that if X satisfies a one-step minorization condition, then pi can be represented as an infinite mixture. The distributions in the mixture are associated with the hitting times on an accessible atom introduced via the splitting construction of Athreya and Ney [Trans. Amer. Math. Soc. 245 (1978) 493-501] and Nummelin [Z. Wahrsch. Verw. Gebiete 43 (1978) 309-318]. When the small set in the minorization condition is the entire state space, our mixture representation of pi reduces to a simple formula, first derived by Breyer and Roberts [Methodol. Comput. Appl. Probab. 3 (2001) 161-177] from which samples can be easily drawn. Despite the fact that the derivation of this formula involves no coupling or backward simulation arguments, the formula can be used to reconstruct perfect sampling algorithms based on coupling from the past (CFTP) such as Murdoch and Green's [Scand. J. Statist. 25 (1998) 483-502] Multigamma Coupler and Wilson's [Random Structures Algorithms 16 (2000) 85-113] Read-Once CFTP algorithm. In the general case where the state space is not necessarily 1-small, under the assumption that X satisfies a geometric drift condition, our mixture representation can be used to construct an arbitrarily accurate approximation to pi from which it is straightforward to sample. One potential application of this approximation is as a starting distribution for a Markov chain Monte Carlo algorithm based on X.