Bayesian calibration with adaptive model discrepancy - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Reports Year : 2022

Bayesian calibration with adaptive model discrepancy

Abstract

We investigate a computer model calibration technique inspired by the wellknown Bayesian framework of Kennedy and O'Hagan. We tackle the full Bayesian formulation where model parameter and model discrepancy hyperparameters are estimated jointly and reduce the problem dimensionality by introducing a functional relationship that we call the Full Maximum a Posteriori (FMP) method. This method also eliminates the need for a true value of model parameters that caused identifiability issues in the KOH formulation. When the joint posterior is approximated as a mixture of Gaussians, the FMP calibration is proved to avoid some pitfalls of the KOH calibration, namely missing some probability regions and underestimating the posterior variance. We then illustrate two numerical examples where both model error and measurement uncertainty are estimated together. Using the solution to the full Bayesian problem as a reference, we show that the FMP results are accurate, robust and avoid the need for high-dimensional Markov Chains for sampling.
Fichier principal
Vignette du fichier
elsarticle-template.pdf (830.34 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-04323574 , version 1 (24-10-2022)
hal-04323574 , version 2 (05-12-2023)

Identifiers

  • HAL Id : hal-04323574 , version 1

Cite

Nicolas Leoni, Olivier Le Maitre, Maria Giovanna Rodio, Pietro Marco Congedo. Bayesian calibration with adaptive model discrepancy. [Research Report] Inria Saclay - Île de France; CEA - Commissariat à l'énergie atomique et aux énergies alternatives; CNRS. 2022. ⟨hal-04323574v1⟩
109 View
274 Download

Share

Gmail Facebook X LinkedIn More