Discrete to continuum limits in Bayesian inverse problems

DOI10.48550/arXiv.2607.27408MaRDI QIDQ7284489arXivFDO

Authors Alexander Katsevich

Publication date 29 July 2026

Published in arXiv

Tldr This paper develops a posterior-level discrete-to-continuum theory for Bayesian inverse problems with finite-dimensional priors and general convex GLM-type likelihoods. The main contribution is proving that in limits of grid refinement and noise vanishing at appropriate rates, MAP estimates from the discretized problem converge to the continuum solution, with fluctuations given by a limiting Gaussian distribution.

Full work available at URL https://doi.org/10.48550/arXiv.2607.27408

arXiv classification

math.NAmath.ST

Cites work

• A path-integral approach to Bayesian inference for inverse problems using the semiclassical approximation
• A PDE construction of the Euclidean \Phi^4_3 quantum field theory
• An introduction to infinite-dimensional analysis
• Approximation of Bayesian Inverse Problems for PDEs
• Bayesian analysis in inverse problems
• Bayesian inverse problems with Gaussian priors
• Bayesian non-linear statistical inverse problems
• Consistent Inversion of Noisy Non-Abelian X-Ray Transforms
• Continuum limits of posteriors in graph Bayesian inverse problems
• Discretization-invariant Bayesian inversion and Besov space priors
• Efficient nonparametric Bayesian inference for X-ray transforms
• Fractional Sobolev Spaces and Functions of Bounded Variation
• Generalized linear models
• High-dimensional Laplace asymptotics up to the concentration threshold
• Improved dimension dependence in the Bernstein von Mises Theorem via a new Laplace approximation bound
• Interacting Langevin diffusions: gradient structure and ensemble Kalman sampler
• Inverse problems: a Bayesian perspective
• Iterative methods for fixed point problems in Hilbert spaces
• Non-Gaussian statistical inverse problems. Part I: Posterior distributions
• On infinite-dimensional hierarchical probability models in statistical inverse problems
• On some information-theoretic aspects of non-linear statistical inverse problems
• Physics-informed information field theory for modeling physical systems with uncertainty quantification
• Quantum field theory and critical phenomena
• Regularity and numerical approximation of fractional elliptic differential equations on compact metric graphs
• Second order ensemble Langevin method for sampling and inverse problems
• The Bayesian Approach To Inverse Problems
• The best constant of discrete Sobolev inequality
• The best constant of discrete Sobolev inequality corresponding to a bending problem of a string
• The best rank-1 approximation of a symmetric tensor and related spherical optimization problems
• Uncertainty Quantification and Weak Approximation of an Elliptic Inverse Problem

This page was built for publication: Discrete to continuum limits in Bayesian inverse problems

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q7284489)