Bayesian analysis of hierarchical multi-fidelity codes. - Archive ouverte HAL Accéder directement au contenu
Autre Publication Scientifique Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences Année : 2012

Bayesian analysis of hierarchical multi-fidelity codes.

Résumé

This paper deals with the Gaussian process based approximation of a code which can be run at different levels of accuracy. This co-kriging method allows us to improve a surrogate model of a complex computer code using fast approximations of it. In particular, we focus on the case of a large number of code levels on the one hand and on a Bayesian approach when we have 2 levels on the other hand. Moreover, based on a Bayes linear formulation, an extension of the universal kriging equations are provided for the co-kriging model. We also address the problem of nested space-filling design for multi-fidelity computer experiments and we provide a significant simplification of the computation of the co-kriging cross-validation equations. A hydrodynamic simulator example is used to illustrate the comparison Bayesian versus non-Bayesian co-kriging. A thermodynamic example is used to illustrate the comparison between 2-level and 3-level co-kriging.
Fichier principal
Vignette du fichier
MainDocument.pdf (325.64 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-00654716 , version 1 (22-12-2011)
hal-00654716 , version 2 (24-09-2012)

Identifiants

Citer

Loic Le Gratiet. Bayesian analysis of hierarchical multi-fidelity codes.. 2012, pp.RSPA-2011-0742. ⟨hal-00654716v1⟩
272 Consultations
362 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More