Well-posed Bayesian inverse problems and heavy-tailed stable quasi-Banach space priors

DOI10.3934/IPI.2017040zbMATH Open1368.65085arXiv1605.05898OpenAlexW2404700421MaRDI QIDQ2013863FDOQ2013863

Authors: Yanyan Li 

Publication date: 9 August 2017

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

Abstract: This article extends the framework of Bayesian inverse problems in infinite-dimensional parameter spaces, as advocated by Stuart (Acta Numer. 19:451--559, 2010) and others, to the case of a heavy-tailed prior measure in the family of stable distributions, such as an infinite-dimensional Cauchy distribution, for which polynomial moments are infinite or undefined. It is shown that analogues of the Karhunen--Lo`eve expansion for square-integrable random variables can be used to sample such measures on quasi-Banach spaces. Furthermore, under weaker regularity assumptions than those used to date, the Bayesian posterior measure is shown to depend Lipschitz continuously in the Hellinger metric upon perturbations of the misfit function and observed data.

Full work available at URL: https://arxiv.org/abs/1605.05898

Recommendations

• Well-posed Bayesian inverse problems with infinitely divisible and heavy-tailed prior measures
• On the well-posedness of Bayesian inverse problems
• Well-Posed Bayesian Inverse Problems: Priors with Exponential Tails
• Besov priors for Bayesian inverse problems
• Quasi-Monte Carlo Bayesian estimation under Besov priors in elliptic inverse problems
• Bayesian inverse problems with non-conjugate priors
• Bayesian approach to inverse problems for functions with a variable-index Besov prior
• Approximation of Bayesian Inverse Problems for PDEs
• Discretization-invariant Bayesian inversion and Besov space priors
• On the local Lipschitz stability of Bayesian inverse problems

zbMATH Keywords

uncertainty quantificationstable distributionwell-posedness[https://portal.mardi4nfdi.de/w/index.php?title=+Special%3ASearch&search=Karhunen-Lo%EF%BF%BD%EF%BF%BDve+expansion&go=Go Karhunen-Lo��ve expansion]heavy-tailed distributionquasi-Banach spacesBayesian inverse problems

Mathematics Subject Classification ID

Infinitely divisible distributions; stable distributions (60E07) Bayesian inference (62F15) Nonparametric robustness (62G35) Probability theory on linear topological spaces (60B11) Inverse problems for PDEs (35R30) Numerical solution to inverse problems in abstract spaces (65J22) Set functions and measures and integrals in infinite-dimensional spaces (Wiener measure, Gaussian measure, etc.) (28C20)

Cites Work

• A Method for Simulating Stable Random Variables
• Statistical and computational inverse problems.
• On the Total Variation and Hellinger Distance Between Signed Measures; an Application to Product Measures
• Title not available (Why is that?)
• Title not available (Why is that?)
• Discretization-invariant Bayesian inversion and Besov space priors
• Inverse problems: a Bayesian perspective
• Can one use total variation prior for edge-preserving Bayesian inversion?
• Infinite dimensional analysis. A hitchhiker's guide.
• Differentiable measures and the Malliavin calculus
• Title not available (Why is that?)
• \(p\)-frames in separable Banach spaces
• Some Properties of Hilbert Scales
• Qualitative Robustness in Bayesian Inference
• Besov priors for Bayesian inverse problems
• Title not available (Why is that?)
• Monte Carlo inference in econometric models with symmetric stable disturbances
• Well-Posed Bayesian Inverse Problems: Priors with Exponential Tails

Cited In (23)

• Sparsity-promoting and edge-preserving maximum a posteriori estimators in non-parametric Bayesian inverse problems
• Hyperpriors for Matérn fields with applications in Bayesian inversion
• Γ-convergence of Onsager–Machlup functionals: II. Infinite product measures on Banach spaces
• A probabilistic finite element method based on random meshes: a posteriori error estimators and Bayesian inverse problems
• Cauchy Markov random field priors for Bayesian inversion
• Cauchy difference priors for edge-preserving Bayesian inversion
• Random time step probabilistic methods for uncertainty quantification in chaotic and geometric numerical integration
• Bayesian Inverse Problems Are Usually Well-Posed
• Two Metropolis--Hastings Algorithms for Posterior Measures with Non-Gaussian Priors in Infinite Dimensions
• Heavy-tailed Bayesian nonparametric adaptation
• Lipschitz continuity of probability kernels in the optimal transport framework
• A Statistical Approach to Surface Metrology for 3D-Printed Stainless Steel
• Bayesian Probabilistic Numerical Methods
• Well-Posed Bayesian Inverse Problems: Priors with Exponential Tails
• An Order-Theoretic Perspective on Modes and Maximum A Posteriori Estimation in Bayesian Inverse Problems
• Bayesian neural network priors for edge-preserving inversion
• 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
• Bayesian Inversion by ω-Complete Cone Duality
• Random Forward Models and Log-Likelihoods in Bayesian Inverse Problems
• On the local Lipschitz stability of Bayesian inverse problems
• Stability of Gibbs Posteriors from the Wasserstein Loss for Bayesian Full Waveform Inversion

This page was built for publication: Well-posed Bayesian inverse problems and heavy-tailed stable quasi-Banach space priors

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2013863)