A supermartingale approach to Gaussian process based sequential design of experiments

Abstract : Gaussian process (GP) models have become a well-established framework for the adaptive design of costly experiments, and notably of computer experiments. GP-based sequential designs have been found practically efficient for various objectives, such as global optimization (estimating the global maximum or maximizer(s) of a function), reliability analysis (estimating a probability of failure) or the estimation of level sets and excursion sets. In this paper, we study the consistency of an important class of sequential designs, known as stepwise uncertainty reduction (SUR) strategies. Our approach relies on the key observation that the sequence of residual uncertainty measures, in SUR strategies, is generally a supermartingale with respect to the filtration generated by the observations. This observation enables us to establish generic consistency results for a broad class of SUR strategies. The consistency of several popular sequential design strategies is then obtained by means of this general result. Notably, we establish the consistency of two SUR strategies proposed by Bect, Ginsbourger, Li, Picheny and Vazquez (Stat. Comp., 2012)—to the best of our knowledge, these are the first proofs of consistency for GP-based sequential design algorithms dedicated to the estimation of excursion sets and their measure. We also establish a new, more general proof of consistency for the expected improvement algorithm for global optimization which, unlike previous results in the literature, applies to any GP with continuous sample paths.
Liste complète des métadonnées

Littérature citée [11 références]  Voir  Masquer  Télécharger

https://hal.archives-ouvertes.fr/hal-01351088
Contributeur : Julien Bect <>
Soumis le : jeudi 13 septembre 2018 - 12:12:15
Dernière modification le : mercredi 12 décembre 2018 - 15:15:33
Document(s) archivé(s) le : vendredi 14 décembre 2018 - 13:23:19

Fichiers

surconv-paper-v4.pdf
Fichiers produits par l'(les) auteur(s)

Licence


Distributed under a Creative Commons Paternité - Pas d'utilisation commerciale - Pas de modification 4.0 International License

Identifiants

  • HAL Id : hal-01351088, version 4
  • ARXIV : 1608.01118

Citation

Julien Bect, François Bachoc, David Ginsbourger. A supermartingale approach to Gaussian process based sequential design of experiments. 2018. 〈hal-01351088v4〉

Partager

Métriques

Consultations de la notice

266

Téléchargements de fichiers

90