BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Inference in non-linear dynamical systems -- a machine learning pe
 rspective - Carl Edward Rasmussen and Andrew McHutchon\, Engineering Depar
 tment\, University of Cambridge
DTSTART:20130715T150000Z
DTEND:20130715T161500Z
UID:TALK46167@talks.cam.ac.uk
CONTACT:Dr Jason Z JIANG
DESCRIPTION:Inference in discrete-time non-linear dynamical systems is oft
 en done using the Extended Kalman Filtering and Smoothing (EKF) algorithm\
 , which provides a Gaussian approximation to the posterior based on local 
 linearisation of the dynamics.  In challenging problems\, when the non-lin
 earities are significant and the signal to noise ratio is poor\, the EKF p
 erforms poorly. In this talk we will discuss an alternative algorithm deve
 loped in the machine learning community which is based message passing in 
 Factor Graphs and the Expectation Propagation (EP) approximation. We will 
 show this method provides a consistent and accurate Gaussian approximation
  to the posterior enabling learning using Expectation Maximisation (EM) ev
 en in cases when the EKF fails.
LOCATION: Cambridge University Engineering Department\, CBL Seminar room B
 E4-38
END:VEVENT
END:VCALENDAR