We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:


Current browse context:


Change to browse by:

References & Citations


(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Statistics > Methodology

Title: Probabilistic Numerical Methods for Partial Differential Equations and Bayesian Inverse Problems

Abstract: This paper develops a probabilistic numerical method for solution of partial differential equations (PDEs) and studies application of that method to PDE-constrained inverse problems. This approach enables the solution of challenging inverse problems whilst accounting, in a statistically principled way, for the impact of discretisation error due to numerical solution of the PDE. In particular, the approach confers robustness to failure of the numerical PDE solver, with statistical inferences driven to be more conservative in the presence of substantial discretisation error. Going further, the problem of choosing a PDE solver is cast as a problem in the Bayesian design of experiments, where the aim is to minimise the impact of solver error on statistical inferences; here the challenge of non-linear PDEs is also considered. The method is applied to parameter inference problems in which discretisation error in non-negligible and must be accounted for in order to reach conclusions that are statistically valid.
Subjects: Methodology (stat.ME); Numerical Analysis (math.NA); Statistics Theory (math.ST); Computation (stat.CO)
Cite as: arXiv:1605.07811 [stat.ME]
  (or arXiv:1605.07811v3 [stat.ME] for this version)

Submission history

From: Jon Cockayne [view email]
[v1] Wed, 25 May 2016 10:22:19 GMT (4744kb,D)
[v2] Mon, 10 Jul 2017 08:55:10 GMT (2833kb,D)
[v3] Tue, 11 Jul 2017 15:22:34 GMT (2833kb,D)

Link back to: arXiv, form interface, contact.