BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Variational Bayesian inference for PDE based inverse problems - Ie
 va Kazlauskaite (University of Cambridge)
DTSTART:20230127T140000Z
DTEND:20230127T150000Z
UID:TALK194893@talks.cam.ac.uk
CONTACT:Qingyuan Zhao
DESCRIPTION:In this talk I will discuss inference in PDE based Bayesian in
 verse problems and present our recent work on variational inference as an 
 alternative to MCMC for this class of problems. In this work\, we propose 
 a family of Gaussian trial distributions parametrised by precision matrice
 s\, taking advantage of the inherent sparsity of the inverse problem encod
 ed in its finite element discretisation.  We utilise stochastic optimisati
 on to efficiently estimate the variational objective and provide an empiri
 cal assessment of the performance. Furthermore\, I will mention some recen
 t work that utilises physics-informed neural network as an alternative to 
 the classical finite element solvers and illustrate how these can be used 
 in PDE based forward and inverse problems.
LOCATION:MR12\, Centre for Mathematical Sciences
END:VEVENT
END:VCALENDAR