BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:The Bayesian Approach To Inverse Problems - Andrew Stuart\, Univer
 sity of Warwick
DTSTART:20130308T160000Z
DTEND:20130308T170000Z
UID:TALK42459@talks.cam.ac.uk
CONTACT:Richard Samworth
DESCRIPTION:Many problems in the physical sciences\nrequire the determinat
 ion of an unknown\nfield from a finite set of indirect measurements.\nExam
 ples include oceanography\, oil recovery\,\nwater resource management and 
 weather forecasting.\nThe Bayesian approach to these problems\nis natural 
 for many reasons\, including the\nunder-determined and ill-posed nature of
  the inversion\,\nthe noise in the data and the uncertainty in\nthe differ
 ential equation models used to describe\ncomplex mutiscale physics.\n\nIn 
 this talk I will describe the advantages of\nformulating Bayesian inversio
 n on function space\nin order to solve these problems. I will overview\nth
 eoretical results concerning well-posedness of\nthe posterior distribution
 \, approximation theorems\nfor the posterior distribution\, and specially\
 nconstructed MCMC methods to explore the posterior\ndistribution when the 
 prior is a Gaussian random field.\nI will also highlight the widespread us
 e by practitioners\nof various ad hoc algorithms such as the Ensemble Kalm
 an Filter\,\nand the need for mathematical and statistical\nanalysis of th
 ese ad hoc algorithms.\n\nIntroductory reading and references may be found
  in:\n\nhttp://arxiv.org/abs/1202.0709 [arxiv.org]\nhttp://arxiv.org/abs/1
 209.2736 [arxiv.org]\n
LOCATION:MR12\, CMS\, Wilberforce Road\, Cambridge\, CB3 0WB
END:VEVENT
END:VCALENDAR