BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Variational inference for partially observed diffusion processes -
  Dr. Cedric Archambeau (University College London)
DTSTART:20080616T130000Z
DTEND:20080616T140000Z
UID:TALK12238@talks.cam.ac.uk
CONTACT:Zoubin Ghahramani
DESCRIPTION:Variational inference is an approximate inference scheme popul
 ar in Machine Learning. In contrast to MCMC it leads to a deterministic so
 lution\, is relatively quick to converge and its convergence can easily be
  checked by monitoring the evolution of the variational bound. In this tal
 k we will recall the basic concepts underlying variational EM and show how
  this framework can be extended to continuous-time stochastic processes. M
 ore specifically\, we will apply the variational approach to partially obs
 erved diffusion processes and discuss parameter inference in this context.
  I will also discuss the links with statistical linearisation and sigma-po
 int transformations.
LOCATION:Engineering Department\, CBL Room 438
END:VEVENT
END:VCALENDAR