Certified dimension reduction in nonlinear Bayesian inverse problems

Olivier Zahm 1 Tiangang Cui 2 Kody Law 3 Alessio Spantini 4 Youssef Marzouk 4
1 AIRSEA - Mathematics and computing applied to oceanic and atmospheric flows
Grenoble INP - Institut polytechnique de Grenoble - Grenoble Institute of Technology, UGA - Université Grenoble Alpes, LJK - Laboratoire Jean Kuntzmann, Inria Grenoble - Rhône-Alpes
Abstract : We propose a dimension reduction technique for Bayesian inverse problems with nonlinear forward operators, non-Gaussian priors, and non-Gaussian observation noise. The likelihood function is approximated by a ridge function, i.e., a map which depends non-trivially only on a few linear combinations of the parameters. We build this ridge approximation by minimizing an upper bound on the Kullback-Leibler divergence between the posterior distribution and its approximation. This bound, obtained via logarithmic Sobolev inequalities, allows one to certify the error of the posterior approximation. Computing the bound requires computing the second moment matrix of the gradient of the log-likelihood function. In practice, a sample-based approximation of the upper bound is then required. We provide an analysis that enables control of the posterior approximation error due to this sampling. Numerical and theoretical comparisons with existing methods illustrate the benefits of the proposed methodology.
Complete list of metadatas

Cited literature [47 references]  Display  Hide  Download

Contributor : Olivier Zahm <>
Submitted on : Tuesday, July 10, 2018 - 10:42:26 AM
Last modification on : Friday, April 19, 2019 - 4:55:18 PM
Long-term archiving on: Thursday, October 11, 2018 - 12:16:13 PM


Files produced by the author(s)


  • HAL Id : hal-01834039, version 1



Olivier Zahm, Tiangang Cui, Kody Law, Alessio Spantini, Youssef Marzouk. Certified dimension reduction in nonlinear Bayesian inverse problems. 2018. ⟨hal-01834039⟩



Record views


Files downloads