Adaptive importance sampling in general mixture classes

被引：180

作者：

Cappe, Olivier ^{[2
]}

Douc, Randal ^{[3
]}

Guillin, Arnaud ^{[4
]}

Marin, Jean-Michel ^{[1
,5
]}

Robert, Christian P. ^{[5
,6
]}

机构：

[1] INRIA Saclay, Project Select, Orsay, France

[2] TELECOM ParisTech, CNRS, LTCI, Paris, France

[3] TELECOM SudParis, Evry, France

[4] Ecole Cent Marseille, CNRS, LATP, Marseille, France

[5] INSEE, CREST, Paris, France

[6] Univ Paris 09, CNRS, CEREMADE, Paris, France

来源：

STATISTICS AND COMPUTING | 2008年 / 18卷 / 04期

关键词：

Importance sampling; Adaptive Monte Carlo; Mixture model; Entropy; Kullback-Leibler divergence; EM algorithm; Population Monte Carlo;

D O I：

10.1007/s11222-008-9059-x

中图分类号：

TP301 [理论、方法];

学科分类号：

081202 ;

摘要：

In this paper, we propose an adaptive algorithm that iteratively updates both the weights and component parameters of a mixture importance sampling density so as to optimise the performance of importance sampling, as measured by an entropy criterion. The method, called M-PMC, is shown to be applicable to a wide class of importance sampling densities, which includes in particular mixtures of multivariate Student t distributions. The performance of the proposed scheme is studied on both artificial and real examples, highlighting in particular the benefit of a novel Rao-Blackwellisation device which can be easily incorporated in the updating scheme.

引用

页码：447 / 459

页数：13

共 14 条

[1]

[Anonymous], 2006, R LANG ENV STAT COMP

[2]

[Anonymous], ESAIM PROBAB STAT

[3]

[Anonymous], 2005, INFERENCE HIDDEN MAR, DOI DOI 10.1007/0-387-28982-8

[4] Population Monte Carlo [J].