Multilevel Markov Chain Monte Carlo for Bayesian Elliptic Inverse Problems with Besov Random Tree Priors

TL;DR

This paper introduces a multilevel Monte Carlo-FEM algorithm for Bayesian elliptic inverse problems using Besov random tree priors, effectively handling non-uniform ellipticity and complex parameter spaces.

Contribution

It develops a novel multilevel Monte Carlo method incorporating Besov random tree priors for elliptic inverse problems, addressing non-uniform ellipticity and complex parameter spaces.

Findings

Algorithm converges in mean-square sense

Achieves essentially optimal asymptotic complexity

Maintains dimension-independent acceptance probabilities

Abstract

We propose a multilevel Monte Carlo-FEM algorithm to solve elliptic Bayesian inverse problems with "Besov random tree prior". These priors are given by a wavelet series with stochastic coefficients, and certain terms in the expansion vanishing at random, according to the law of so-called Galton-Watson trees. This allows to incorporate random fractal structures and large deviations in the log-diffusion, which occur naturally in many applications from geophysics or medical imaging. This framework entails two main difficulties: First, the associated diffusion coefficient does not satisfy a uniform ellipticity condition, which leads to non-integrable terms and thus divergence of standard multilevel estimators. Secondly, the associated space of parameters is Polish, but not a normed linear space. We address the first point by introducing cut-off functions in the estimator to compensate for the non-integrable terms, while the second issue is resolved by employing an independence Metropolis-Hastings sampler. The resulting algorithm converges in the mean-square sense with essentially optimal asymptotic complexity, and dimension-independent acceptance probabilities.