Sequential design of Gaussian process surrogates using pre-posterior analysis and Bayesian model averaging - CEA - Commissariat à l’énergie atomique et aux énergies alternatives Accéder directement au contenu
Communication Dans Un Congrès Année : 2023

Sequential design of Gaussian process surrogates using pre-posterior analysis and Bayesian model averaging

W. Fauriat
  • Fonction : Auteur

Résumé

To build a surrogate model from experimental data-from tests or computer simulations-numerous options may exist when choosing a mathematical form. This is true for Gaussian Processes (GP) models, which may or may not include a regression basis-or mean trend-and be built on different correlation structures-through the selected covariance kernel. When data is scarce and prior information on the modeled phenomena is poor, it may be difficult to come to a conclusion on important decisions such as model selection or model improvement. If additional experimental information can be collected, often at significant cost, it is interesting to carry out model selection sequentially and efficiently. In this paper, we propose to leverage the ability of GPs to provide probabilistic descriptions and use it to look for the next "best" point in the design space using a pre-posterior analysis scheme, or Value Of Information (VoI) evaluation. At such point, we expect to get the most relevant information, when the aim is to reduce expected prediction error, given a previous state of knowledge on the likelihood of various modeling options, e.g. using the idea of Bayesian Model Averaging (BMA). With successive queried points, we update our respective beliefs in these options through an "information-optimal" exploration of the design space-given current expectations according to priors. Hence, we attempt to learn efficiently both model structure and parameters.
Fichier principal
Vignette du fichier
WF_ICASP14paper_vfinal.pdf (689.97 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

cea-04170299 , version 1 (25-07-2023)

Identifiants

  • HAL Id : cea-04170299 , version 1

Citer

W. Fauriat. Sequential design of Gaussian process surrogates using pre-posterior analysis and Bayesian model averaging. ICASP 14 - 14th International Conference on Applications of Statistics and Probability in Civil Engineering, Jul 2023, Dublin, Ireland. ⟨cea-04170299⟩

Collections

CEA DAM
27 Consultations
33 Téléchargements

Partager

Gmail Facebook X LinkedIn More