Sampling from a log-concave distribution with compact support with proximal Langevin Monte Carlo

Abstract : This paper presents a detailed theoretical analysis of the Langevin Monte Carlo sampling algorithm recently introduced in [DMP16] when applied to log-concave probability distributions that are restricted to a convex body K. This method relies on a regularisation procedure involving the Moreau-Yosida envelope of the indicator function associated with K. Explicit convergence bounds in total variation norm and in Wasserstein distance of order 1 are established. In particular, we show that the complexity of this algorithm given a first order oracle is polynomial in the dimension of the state space. Finally, some numerical experiments are presented to compare our method with competing MCMC approaches from the literature.
Type de document :
Article dans une revue
Proceedings of Machine Learning Research, PMLR, 2017, 65, pp.319-342
Liste complète des métadonnées

Littérature citée [30 références]  Voir  Masquer  Télécharger

https://hal.inria.fr/hal-01648665
Contributeur : Nicolas Brosse <>
Soumis le : lundi 27 novembre 2017 - 23:27:53
Dernière modification le : jeudi 12 avril 2018 - 01:50:58

Fichier

Identifiants

  • HAL Id : hal-01648665, version 1

Citation

Nicolas Brosse, Alain Durmus, Eric Moulines, Marcelo Pereyra. Sampling from a log-concave distribution with compact support with proximal Langevin Monte Carlo. Proceedings of Machine Learning Research, PMLR, 2017, 65, pp.319-342. 〈hal-01648665〉

Partager

Métriques

Consultations de la notice

160

Téléchargements de fichiers

17