The Bayesian Approach to Inverse Problems

By Masoumeh Dashti, Andrew M. Stuart. 2013

675 citations recorded by the source corpus.

These lecture notes highlight the mathematical and computational structure relating to the formulation of, and development of algorithms for, the Bayesian approach to inverse problems in differential equations. This approach is fundamental in the quantification of uncertainty within applications in volving the blending of mathematical models with data. The finite dimensional situation is described first, along with some motivational examples. Then the development of probability measures on separable Banach space is undertaken, using a random series over an infinite set of functions to construct draws; these probability measures are used as priors in the Bayesian approach to inverse problems. Regularity of draws from the priors is studied in the natural Sobolev or Besov spaces implied by the choice of functions in the random series construction, and the Kolmogorov continuity theorem is used to extend regularity considerations to the space of Holder continuous functions. Bayes’ theorem is de rived in this prior setting, and here interpreted as finding conditions under which the posterior is absolutely continuous with respect to the prior, and determining a formula for the Radon-Nikodym derivative in terms of the likelihood of the data. Having established the form of the posterior, we then describe various properties common to it in the infinite dimensional setting. These properties include well-posedness, approximation theory, and the existence of maximum a posteriori estimators. We then describe measure-preserving dynamics, again on the infinite dimensional space, including Markov chain-Monte C arlo and sequential Monte Carlo methods, and measure-preserving reversible stochastic differential equations. By formulating the theory and algorithms on the underlying infinite dimensional space, we obtain a framework suitable for rigorous analysis of the accuracy of reconstructions, of computational complexity, as well as naturally constructing algorithms which perform well under mesh refinement, since they are inherently well-defined in infinite dimensions.

Subfields: Numerical Analysis, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/a4ef807ab0f3efe64e0c08a38ba03aea9f0d0837

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Numerical Analysis

In Statistics and Probability

Citations in the corpus

Empirical Bayesian regularization of the inverse acoustic problem (2015, background): “A general reference on the Bayesian approach t o inverse problems is [46].”
Bayesian Probabilistic Numerical Methods (2017, background): “Stuart [2010], Dashti and Stuart [2013], and Sullivan [2016].”
Gaussian Process Behaviour in Wide Deep Neural Networks (2018, methodology): “This metric metrises the product topology of the product of countably many copies of R with the usual Euclidean topology (Dashti and Stuart, 2013).”
Solving inverse problems using data-driven models (2019, methodology): “Our brief survey in Sections 3.1, 3.2.1 and 3.2.2 is based on the excellent survey papers by Evans and Stark (2002), Stuart (2010) and Dashti and Stuart (2017).”
Probabilistic Meshless Methods for Partial Differential Equations and Bayesian Inverse Problems (2016, methodology): “9 of Dashti and Stuart [2015], it is sufficient to establish that the two potentials Φh and Φ are asymptotically identical.”
Bilevel approaches for learning of variational imaging models (2015, background): “Appropriate data ﬁdelity distances d are mostly driven by statistical considerations that model our knowledge of the data distribution [47, 49].”
Sparse-grid, reduced-basis Bayesian inversion (2015, background): “In Section 2, we present the problem of Bayesian inversion of parametric operator equations in a function space setting, based on [55, 22], and for the general class of affine-parametric operator equations considered here.”