On the local Lipschitz stability of Bayesian inverse problems

DOI10.1088/1361-6420/AB6F43zbMATH Open1469.62216arXiv1906.07120OpenAlexW3103335497MaRDI QIDQ5000612FDOQ5000612

Authors: Björn Sprungk 

Publication date: 14 July 2021

Published in: Inverse Problems (Search for Journal in Brave)

Abstract: In this note we consider the stability of posterior measures occuring in Bayesian inference w.r.t. perturbations of the prior measure and the log-likelihood function. This extends the well-posedness analysis of Bayesian inverse problems. In particular, we prove a general local Lipschitz continuous dependence of the posterior on the prior and the log-likelihood w.r.t. various common distances of probability measures. These include the total variation, Hellinger, and Wasserstein distance and the Kullback-Leibler divergence. We only assume the boundedness of the likelihoods and measure their perturbations in an $L^p$-norm w.r.t. the prior. The obtained stability yields under mild assumptions the well-posedness of Bayesian inverse problems, in particular, a well-posedness w.r.t. the Wasserstein distance. Moreover, our results indicate an increasing sensitivity of Bayesian inference as the posterior becomes more concentrated, e.g., due to more or more accurate data. This confirms and extends previous observations made in the sensitivity analysis of Bayesian inference.

Full work available at URL: https://arxiv.org/abs/1906.07120

Recommendations

• On the well-posedness of Bayesian inverse problems
• Well-Posed Bayesian Inverse Problems: Priors with Exponential Tails
• Well-posed Bayesian inverse problems and heavy-tailed stable quasi-Banach space priors
• Stability and infinitesimal robustness of posterior distributions and posterior quantities
• Posterior consistency for Bayesian inverse problems through stability and regression results

zbMATH Keywords

Bayesian inferencerobust statisticsWasserstein distanceKullback-Leibler divergenceinverse problemswell-posednessHellinger distance

Mathematics Subject Classification ID

Convergence of probability measures (60B10) Bayesian inference (62F15)

Cites Work

• Gaussian measures in Banach spaces
• Statistical and computational inverse problems.
• Title not available (Why is that?)
• Asymptotic Statistics
• Whittle-Matérn priors for Bayesian statistical inversion with applications in electrical impedance tomography
• On Choosing and Bounding Probability Metrics
• Optimal Transport
• On the consistency of Bayes estimates
• Nonparametric Bernstein-von Mises theorems in Gaussian white noise
• On the Bernstein-von Mises phenomenon for nonparametric Bayes procedures
• Robust Bayesian analysis
• An overview of robust Bayesian analysis. (With discussion)
• On the Bernstein-von Mises theorem with infinite-dimensional parameters
• Robust Bayesian analysis: sensitivity to the prior
• Inverse problems: a Bayesian perspective
• Robust Bayesian analysis using divergence measures
• Perturbation theory for Markov chains via Wasserstein distance
• Local sensitivity diagnostics for Bayesian inference
• Title not available (Why is that?)
• Title not available (Why is that?)
• The Bernstein-von Mises theorem under misspecification
• On the Bernstein-von Mises phenomenon in the Gaussian white noise model
• On a Formula for the L2 Wasserstein Metric between Measures on Euclidean and Hilbert Spaces
• Title not available (Why is that?)
• A new metric for probability distributions
• Asymptotic coupling and a general form of Harris' theorem with applications to stochastic delay equations
• Laplace's method revisited: Weak convergence of probability measures
• Title not available (Why is that?)
• Well-posed Bayesian inverse problems and heavy-tailed stable quasi-Banach space priors
• Brittleness of Bayesian inference under finite information in a continuous world
• On the brittleness of Bayesian inference
• Brittleness of Bayesian inference and new Selberg formulas
• Posterior consistency for Gaussian process approximations of Bayesian posterior distributions
• The dilation phenomenon in robust Bayesian inference. (With discussion)
• Hierarchical Bayesian level set inversion
• Title not available (Why is that?)
• Perturbation bounds for Monte Carlo within metropolis via restricted approximations
• Well-Posed Bayesian Inverse Problems: Priors with Exponential Tails
• Well-Posed Bayesian Inverse Problems with Infinitely Divisible and Heavy-Tailed Prior Measures

Cited In (23)

• A MCMC method based on surrogate model and Gaussian process parameterization for infinite Bayesian PDE inversion
• NF-ULA: normalizing flow-based unadjusted Langevin algorithm for imaging inverse problems
• On global normal linear approximations for nonlinear Bayesian inverse problems
• Optimal experimental design for infinite-dimensional Bayesian inverse problems governed by PDEs: a review
• Well-posed Bayesian inverse problems and heavy-tailed stable quasi-Banach space priors
• Choosing observation operators to mitigate model error in Bayesian inverse problems
• Wasserstein Sensitivity of Risk and Uncertainty Propagation
• Optimal experimental design: formulations and computations
• Stability of doubly-intractable distributions
• Markov chain generative adversarial neural networks for solving Bayesian inverse problems in physics applications
• Brittleness of Bayesian inference under finite information in a continuous world
• Bayesian Inverse Problems Are Usually Well-Posed
• Error Bounds for Some Approximate Posterior Measures in Bayesian Inference
• Image Denoising: The Deep Learning Revolution and Beyond—A Survey Paper
• A Bayesian scheme for reconstructing obstacles in acoustic waveguides
• Stability estimates for the expected utility in Bayesian optimal experimental design
• On the Well-posedness of Bayesian Inverse Problems
• Spectral gaps and error estimates for infinite-dimensional Metropolis-Hastings with non-Gaussian priors
• Γ -convergence of Onsager–Machlup functionals: I. With applications to maximum a posteriori estimation in Bayesian inverse problems
• Learning from small data sets: patch-based regularizers in inverse problems for image reconstruction
• Robustness and exploration of variational and machine learning approaches to inverse problems: an overview
• On Bayesian data assimilation for PDEs with ill-posed forward problems
• Analysis of a Computational Framework for Bayesian Inverse Problems: Ensemble Kalman Updates and MAP Estimators under Mesh Refinement

This page was built for publication: On the local Lipschitz stability of Bayesian inverse problems

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5000612)