BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Approximate Smoothing and Parameter Estimation in High-Dimensional
  State-Space Models - Dr Axel Finke\, CUED
DTSTART:20160519T140000Z
DTEND:20160519T150000Z
UID:TALK65851@talks.cam.ac.uk
CONTACT:Prof. Ramji Venkataramanan
DESCRIPTION:We present an approximate algorithm for estimating additive sm
 oothing functionals in a class of high-dimensional state-space models via 
 sequential Monte Carlo methods. In such high-dimensional settings\, a proh
 ibitively large number of particles\, i.e. growing exponentially in the di
 mension of the state space\, is usually required to obtain useful estimate
 s of such smoothed quantities. Exploiting spatial ergodicity properties of
  the model\, we circumvent this problem via a blocking strategy which lead
 s to approximations that can be computed recursively in time and in parall
 el in space. In particular\, our method enables us to perform maximum-like
 lihood estimation via stochastic gradient-ascent and stochastic EM algorit
 hms. We demonstrate the method on a high-dimensional state-space model. \n
 \nThis is joint work with Sumeetpal S. Singh.
LOCATION:LR6\, Department of Engineering
END:VEVENT
END:VCALENDAR