BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Particle Flow for Near-perfect Sampling in Static and Dynamic Cont
 exts - Prof Simon Maskell\, Liverpool
DTSTART:20150619T143000Z
DTEND:20150619T153000Z
UID:TALK59816@talks.cam.ac.uk
CONTACT:Karthik Tadinada
DESCRIPTION:Particle flow has been described in the literature as a very e
 fficient particle filter with no need for resampling. It has been document
 ed as achieving impressive results against state-of-the-art baselines when
  applied to high dimensional dynamic inference problems. However\, the exi
 sting literature makes it difficult to understand how\, why and when parti
 cle flow works. This talk will put particle flow on a solid theoretical fo
 oting. Links to Langevin diffusions\, marginal particle filters\, SMC samp
 lers and data augmentation will be highlighted. The conditions that result
  in no resampling being necessary will be made clear. To highlight the ben
 efits\, results will demonstrate the efficiency of the technique in both d
 ynamic contexts and static contexts (where particle flow offers untapped p
 otential to outperform\, for example\, MCMC and SMC samplers). 
LOCATION:Computer Laboratory\, William Gates Building\, Lecture Theatre 1
END:VEVENT
END:VCALENDAR