Bayesian inverse problems with partial observations

DOI10.1016/J.TRMI.2018.09.002MaRDI QIDQ2317094zbMATH OpenOpenAlexFDO

Authors Shota Gugushvili, Dong Yan, Aad van der Vaart

Publication date 8 August 2019

Published in Transactions of A. Razmadze Mathematical Institute (Search for Journal in Brave)

Copyright license Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International

Full work available at URL https://arxiv.org/abs/1802.08993

zbMATH Keywords

credible setfrequentist coverageGaussian priorGaussian sequence modelheat equationinverse problemnonparametric Bayesian estimationposterior contraction ratesingular value decompositionVolterra operator

Mathematics Subject Classification ID

Heat equation (35K05) Fourier series and coefficients in several variables (42B05) Gaussian processes (60G15) Density estimation (62G07)

Abstract: We study a nonparametric Bayesian approach to linear inverse problems under discrete observations. We use the discrete Fourier transform to convert our model into a truncated Gaussian sequence model, that is closely related to the classical Gaussian sequence model. Upon placing the truncated series prior on the unknown parameter, we show that as the number of observations $nightarrowinfty,$ the corresponding posterior distribution contracts around the true parameter at a rate depending on the smoothness of the true parameter and the prior, and the ill-posedness degree of the problem. Correct combinations of these values lead to optimal posterior contraction rates (up to logarithmic factors). Similarly, the frequentist coverage of Bayesian credible sets is shown to be dependent on a combination of smoothness of the true parameter and the prior, and the ill-posedness of the problem. Oversmoothing priors lead to zero coverage, while undersmoothing priors produce highly conservative results. Finally, we illustrate our theoretical results by numerical examples.

Recommendations

• Bayesian inverse problems with Gaussian priors
• Posterior contraction rates for the Bayesian approach to linear ill-posed inverse problems
• Bayesian inverse problems with non-conjugate priors
• Bayesian posterior contraction rates for linear severely ill-posed inverse problems
• Bayesian inverse problems with unknown operators

Cites work

• Adaptive Wavelet Galerkin Methods for Linear Inverse Problems
• An introduction to the mathematical theory of inverse problems
• Bayesian inverse problems with Gaussian priors
• Bayesian recovery of the initial condition for the heat equation
• Confidence bands for inverse regression models
• Convergence Rates of General Regularization Methods for Statistical Inverse Problems and Applications
• Convergence rates of posterior distributions.
• Functional analysis. An elementary introduction
• Fundamentals of nonparametric Bayesian inference
• Generalized Discrete Fourier Transform With Nonlinear Phase
• scientific article; zbMATH DE number 47995 (Why is no real title available?)
• scientific article; zbMATH DE number 1197366 (Why is no real title available?)
• Inverse problems and high-dimensional estimation. Stats in the Château summer school, Paris, France, August 31 -- September 4, 2009.
• Model checks in inverse regression models with convolution-type operators
• Nonlinear solution of linear inverse problems by wavelet-vaguelette decomposition
• Nonparametric statistical inverse problems
• Nonstationary inverse problems and state estimation
• Practical Approximate Solutions to Linear Operator Equations When the Data are Noisy
• Sharp adaptation for inverse problems with random noise
• Simultaneous sharp estimation of functions and their derivatives
• The mathematics of computerized tomography

Cited in

(21)

• Estimating anomalies from indirect observations
• Bayesian posterior contraction rates for linear severely ill-posed inverse problems
• Designing truncated priors for direct and inverse Bayesian problems
• Hyperparameter estimation in Bayesian MAP estimation: parameterizations and consistency
• Large deviations for infinite weighted sums of stretched exponential random variables
• On Bayesian consistency for flows observed through a passive scalar
• scientific article; zbMATH DE number 5951039 (Why is no real title available?)
• scientific article; zbMATH DE number 5168223 (Why is no real title available?)
• Besov priors for Bayesian inverse problems
• Non-Gaussian statistical inverse problems. Part I: Posterior distributions
• Bayesian inverse problems with unknown operators
• Bayesian recovery of the initial condition for the heat equation
• scientific article; zbMATH DE number 1415983 (Why is no real title available?)
• Bernstein-von Mises theorems and uncertainty quantification for linear inverse problems
• Bayesian inverse problems with non-commuting operators
• Solving inverse problems using data-driven models
• Transform-based particle filtering for elliptic Bayesian inverse problems
• On empirical Bayes approach to inverse problems
• Choosing observation operators to mitigate model error in Bayesian inverse problems
• Bayesian inverse problems with Gaussian priors
• Bayesian linear inverse problems in regularity scales with discrete observations

This page was built for publication: Bayesian inverse problems with partial observations

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2317094)