BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Particle Learning - Nicholas Polson University of Chicago Booth Sc
 hool of Business
DTSTART:20090515T150000Z
DTEND:20090515T160000Z
UID:TALK18117@talks.cam.ac.uk
CONTACT:8047
DESCRIPTION: This talk will introduce novel particle learning (PL)\nmethod
 s for sequential filtering\,\nparameter learning and smoothing in a genera
 l class of state space\nmodels. The approach extends existing particle met
 hods by\nincorporating unknown fixed parameters\, utilizing sufficient\nst
 atistics\, for the parameters and/or the states\, and allowing for\nnonlin
 earities in the model. We also show how to solve the state\nsmoothing prob
 lem\, integrating out parameter uncertainty. We show that\nour algorithms 
 outperform MCMC\, as well as existing particle filtering\nalgorithms.\n\n
LOCATION:MR12\, CMS\, Wilberforce Road\, Cambridge\, CB3 0WB
END:VEVENT
END:VCALENDAR