ABC likelihood-free methods for model choice in Gibbs random fields
Grelaud, Aude; Robert, Christian P.; Marin, Jean-Michel; Rodolphe, François; Taly, Jean-François (2009), ABC likelihood-free methods for model choice in Gibbs random fields, Bayesian Analysis, 4, 2, p. 317-336. http://dx.doi.org/10.1214/09-BA412
Type
Article accepté pour publication ou publiéLien vers un document non conservé dans cette base
http://arxiv.org/abs/0807.2767v3Date
2009Nom de la revue
Bayesian AnalysisVolume
4Numéro
2Éditeur
International Society for Bayesian Analysis
Pages
317-336
Identifiant publication
Métadonnées
Afficher la notice complèteAuteur(s)
Grelaud, AudeRobert, Christian P.
Marin, Jean-Michel

Rodolphe, François
Taly, Jean-François
Résumé (EN)
Gibbs random fields are polymorphous statistical models that can be used to analyse different types of dependence, in particular for spatially correlated data. However, when those models are faced with the challenge of selecting a dependence structure from many, the use of standard model choice methods is hampered by the unavailability of the normalising constant in the Gibbs likelihood. In particular, from a Bayesian perspective, the computation of the posterior probabilities of the models under competition requires special likelihood-free simulation techniques like the Approximate Bayesian Computation (ABC) algorithm that is intensively used in population Genetics. We show in this paper how to implement an ABC algorithm geared towards model choice in the general setting of Gibbs random fields, demonstrating in particular that there exists a sufficient statistic across models. The accuracy of the approximation to the posterior probabilities can be further improved by importance sampling on the distribution of the models. The practical aspects of the method are detailed through two applications, the test of an iid Bernoulli model versus a first-order Markov chain, and the choice of a folding structure for a protein of Thermotoga maritima implicated into signal transduction processes.Mots-clés
Statistical models; MarkovPublications associées
Affichage des éléments liés par titre et auteur.
-
Grelaud, Aude; Robert, Christian P.; Marin, Jean-Michel (2009) Article accepté pour publication ou publié
-
Rodolphe, François; Marin, Jean-Michel; Taly, Jean-François; Grelaud, Aude; Robert, Christian P. (2009) Communication / Conférence
-
Pudlo, Pierre; Marin, Jean-Michel; Estoup, Arnaud; Cornuet, Jean-Marie; Gautier, Mathieu; Robert, Christian P. (2016) Article accepté pour publication ou publié
-
Marin, Jean-Michel; Pudlo, Pierre; Estoup, Arnaud; Robert, Christian P. (2018) Chapitre d'ouvrage
-
Pillai, Natesh S.; Marin, Jean-Michel; Robert, Christian P. (2011-01) Document de travail / Working paper