Generalized Bayes Approach to Inverse Problems with Model Misspecification

Abstract

We propose a general framework for obtaining probabilistic solutions to PDE-based inverse problems. Bayesian methods are attractive for uncertainty quantification but assume knowledge of the likelihood model or data generation process. This assumption is difficult to justify in many inverse problems, where the specification of the data generation process is not obvious. We adopt a Gibbs posterior framework that directly posits a regularized variational problem on the space of probability distributions of the parameter. We propose a novel model comparison framework that evaluates the optimality of a given loss based on its "predictive performance". We provide cross-validation procedures to calibrate the regularization parameter of the variational objective and compare multiple loss functions. Some novel theoretical properties of Gibbs posteriors are also presented. We illustrate the utility of our framework via a simulated example, motivated by dispersion-based wave models used to characterize arterial vessels in ultrasound vibrometry.

arXiv comment26 pages, 6 figures. Replaced for agreement with the final published version

References in corpus (17)

• Practical Bayesian model evaluation using leave-one-out cross-validation and WAIC
• A General Framework for Updating Belief Distributions
• Bayesian inverse problems with Gaussian priors
• On the Bernstein-von Mises phenomenon for nonparametric Bayes procedures
• Gibbs posterior for variable selection in high-dimensional classification and data mining
• On Russian Roulette Estimates for Bayesian Inference with Doubly-Intractable Likelihoods
• Calibrating general posterior credible regions
• Nonparametric Bernstein-von Mises theorems in Gaussian white noise
• Empirical Bayes posterior concentration in sparse high-dimensional linear models
• Fast Rates for General Unbounded Loss Functions: from ERM to Generalized Bayes
• Brittleness of Bayesian Inference Under Finite Information in a Continuous World
• Gibbs posterior concentration rates under sub-exponential type losses
• Robust leave-one-out cross-validation for high-dimensional Bayesian models
• Gradient-Based Estimation of Uncertain Parameters for Elliptic Partial Differential Equations
• Variational Bayesian strategies for high-dimensional, stochastic design problems
• Towards Low-Photon Nanoscale Imaging: Holographic Phase Retrieval via Maximum Likelihood Optimization
• Adaptive particle-based approximations of the Gibbs posterior for inverse problems