BEGIN:VCALENDAR
PRODID:-//Columba Systems Ltd//NONSGML CPNG/SpringViewer/ICal Output/3.3-
M3//EN
VERSION:2.0
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VEVENT
DTSTAMP:20200220T165507Z
DTSTART:20200310T120000Z
DTEND:20200310T130000Z
SUMMARY:Richard Nickl - Statistical guarantees for the Bayesian approach
to inverse problems
UID:{http://www.columbasystems.com/customers/uom/gpp/eventid/}o7u-k5cftf6
y-kkhw5d
DESCRIPTION:***Postponed due to strike action*** The seminar should be re
scheduled for the autumn.\n\nAbstract: Bayes methods for inverse problem
s have become very popular in applied mathematics in the last decade aft
er seminal work by Persi Diaconis and Andrew Stuart. They provide recons
truction algorithms as well as in-built â€śuncertainty quantificationâ€ť via
Bayesian credible sets\, and particularly for Gaussian process priors c
an be efficiently implemented by MCMC methodology. For linear inverse pr
oblems\, they are closely related to classical penalised least squares m
ethods and thus not fundamentally new\, but for non-linear and non-conve
x problems\, they give genuinely distinct and computable algorithmic alt
ernatives that cannot be studied by variational analysis or convex optim
isation techniques. In this talk we will discuss recent progress in Baye
sian non-parametric statistics that allows to give rigorous statistical
guarantees for posterior consistency in such models\, and illustrate the
theory in a variety of concrete non-linear inverse problems arising wit
h partial differential equations and X-ray transforms.
STATUS:TENTATIVE
TRANSP:TRANSPARENT
CLASS:PUBLIC
LOCATION:Frank Adams 1\, Alan Turing Building\, Manchester
END:VEVENT
END:VCALENDAR