BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Exclusive Pólya Urns and their applications - Christian Steinruec
 ken (University of Cambridge)
DTSTART:20110916T100000Z
DTEND:20110916T110000Z
UID:TALK33035@talks.cam.ac.uk
CONTACT:Zoubin Ghahramani
DESCRIPTION:The Dirichlet Process (DP) and its variants have many nice pro
 perties which make them popular tools in the machine learning community.  
 For some applications\, however\, these processes lack certain desirable f
 eatures\, such as support for fast and exact inference\, or being easy to 
 interface to an arithmetic coder.  Common deployment typically involves ap
 proximate inference methods\, which either distort the true posterior dist
 ribution or come at a high computational cost.  I will show that by droppi
 ng some of the "nice" mathematical properties of the DP\, we can instead c
 onstruct a novel kind of stochastic process which allows exact Bayesian in
 ference\, is fast\, easy to implement\, and can in many cases be used as a
  drop-in replacement for Chinese Restaurant Processes and Pitman-Yor proce
 sses.  I will explain how the new process differs from models in the exist
 ing literature\, and how it can be applied to interesting tasks\, includin
 g hierarchical sequence modelling and data compression.
LOCATION:Engineering Department\, CBL Room 438
END:VEVENT
END:VCALENDAR