BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Butterfly resampling - convergence and central limit theorems for 
 particle filters with constrained interactions - Dr Kari Heine\, Universit
 y of Bristol
DTSTART:20141120T140000Z
DTEND:20141120T150000Z
UID:TALK56121@talks.cam.ac.uk
CONTACT:Fredrik Lindsten
DESCRIPTION:We describe a novel class of particle filters that generalizes
  the classical bootstrap filter in a manner of introducing constraints on 
 the interaction pattern of the particles. In some instances\, the conditio
 nal independence structure of the new algorithm can be expressed as a grap
 h with the same structure as the butterfly diagram of the Cooley-Tukey fas
 t Fourier transform. The main motivation for the interest in these algorit
 hms with sparse independence structure is to lay rigorous foundations for 
 the design of algorithms better suited to modern computing architectures.\
 n\nThe law of large numbers and the central limit theorem (CLT) are establ
 ished for specific instances of the new particle filters. It turns out\, t
 hat the price to pay for the sparseness of the conditional independence st
 ructure is increased asymptotic variance in the CLT\, and\, in some cases\
 , slower rate of convergence that manifests itself as a non-standard scali
 ng in the CLT.
LOCATION:Board Room\, CUED
END:VEVENT
END:VCALENDAR