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

Mathematics > Statistics Theory

Title: Posterior distribution existence and error control in Banach spaces in the Bayesian approach to UQ in inverse problems

Abstract: We generalize the results of \cite{Capistran2016} on expected Bayes factors (BF) to control the numerical error in the posterior distribution to an infinite dimensional setting when considering Banach functional spaces and now in a prior setting. The main result is a bound on the absolute global error to be tolerated by the Forward Map numerical solver, to keep the BF of the numerical vs. the theoretical model near to 1, now in this more general setting, possibly including a truncated, finite dimensional approximate prior measure. In so doing we found a far more general setting to define and prove existence of the infinite dimensional posterior distribution than that depicted in, for example, \cite{Stuart2010}. Discretization consistency and rates of convergence are also investigated in this general setting for the Bayesian inverse problem.
Comments: 42 pages, 6 figures
Subjects: Statistics Theory (math.ST)
MSC classes: 62, 65
Cite as: arXiv:1712.03299 [math.ST]
  (or arXiv:1712.03299v3 [math.ST] for this version)

Submission history

From: Andres Christen [view email]
[v1] Fri, 8 Dec 2017 22:30:01 GMT (23kb)
[v2] Thu, 11 Oct 2018 15:49:46 GMT (385kb)
[v3] Fri, 12 Oct 2018 16:29:37 GMT (385kb)

Link back to: arXiv, form interface, contact.