Convergence of the equi-energy sampler
ESAIM. Proceedings, Tome 19 (2007), pp. 1-5
Cet article a éte moissonné depuis la source EDP Sciences
In a recent paper 'The equi-energy sampler with applications statistical inference and statistical mechanics' Kou, Zhou Wong (Ann. Statist., 34, 1581-1619, 2006.) have presented a new stochastic simulation method called the equi-energy (EE) sampler. This technique is designed to simulate from a probability measure π, perhaps only known up to a normalizing constant. The authors demonstrate that the sampler performs well in quite challenging problems but their convergence results (Theorem 2) appear incomplete. This was pointed out, in the discussion of the paper, by Atchadé Liu (Ann. Statist., 34, 1620-1628, 2006.) who proposed an alternative convergence proof. However, this alternative proof, whilst theoretically correct, does not correspond to the algorithm that is implemented. In this note we consider the difficulties of the proofs as well as pointing out an alternative convergence result established by the authors (Andrieu et al. 2007b).
Affiliations des auteurs :
Christophe Andrieu 1 ; Ajay Jasra 2 ; Arnaud Doucet 3 ; Pierre Del Moral 4
@article{EP_2007_19_a1,
author = {Christophe Andrieu and Ajay Jasra and Arnaud Doucet and Pierre Del Moral},
title = {Convergence of the equi-energy sampler},
journal = {ESAIM. Proceedings},
pages = {1--5},
year = {2007},
volume = {19},
doi = {10.1051/proc:071901},
language = {en},
url = {http://geodesic.mathdoc.fr/articles/10.1051/proc:071901/}
}
TY - JOUR AU - Christophe Andrieu AU - Ajay Jasra AU - Arnaud Doucet AU - Pierre Del Moral TI - Convergence of the equi-energy sampler JO - ESAIM. Proceedings PY - 2007 SP - 1 EP - 5 VL - 19 UR - http://geodesic.mathdoc.fr/articles/10.1051/proc:071901/ DO - 10.1051/proc:071901 LA - en ID - EP_2007_19_a1 ER -
Christophe Andrieu; Ajay Jasra; Arnaud Doucet; Pierre Del Moral. Convergence of the equi-energy sampler. ESAIM. Proceedings, Tome 19 (2007), pp. 1-5. doi: 10.1051/proc:071901
Cité par Sources :