Bernstein--von Mises Theorems and Uncertainty Quantification for Linear Inverse Problems

Matteo Giordano, Hanne Kekkonen

Research output: Contribution to journalArticleScientificpeer-review

Abstract

We consider the statistical inverse problem of recovering an unknown function f from a linear measurement corrupted by additive Gaussian white noise. We employ a nonparametric Bayesian approach with standard Gaussian priors, for which the posterior-based reconstruction of f corresponds to a Tikhonov regulariser f¯ with a reproducing kernel Hilbert space norm penalty. We prove a semiparametric Bernstein-von Mises theorem for a large collection of linear functionals of f, implying that semiparametric posterior estimation and uncertainty quantification are valid and optimal from a frequentist point of view. The result is applied to study three concrete examples that cover both the mildly and severely ill-posed cases: specifically, an elliptic inverse problem, an elliptic boundary value problem and the heat equation. For the elliptic boundary value problem, we also obtain a nonparametric version of the theorem that entails the convergence of the posterior distribution to a prior-independent infinite-dimensional Gaussian probability measure with minimal covariance. As a consequence, it follows that the Tikhonov regulariser f¯ is an efficient estimator of f, and we derive frequentist guarantees for certain credible balls centred at f¯.
Original languageEnglish
Pages (from-to)342-373
Number of pages32
JournalSIAM-ASA Journal on Uncertainty Quantification
Volume8
Issue number1
DOIs
Publication statusPublished - 2020
Externally publishedYes

Keywords

  • Bernstein–von Mises theorems
  • Gaussian priors
  • Tikhonov regularisers
  • Asymptotics of nonparametric Bayes procedures
  • Elliptic partial differential equati

Fingerprint

Dive into the research topics of 'Bernstein--von Mises Theorems and Uncertainty Quantification for Linear Inverse Problems'. Together they form a unique fingerprint.

Cite this