BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//talks.cam.ac.uk//v3//EN
BEGIN:VTIMEZONE
TZID:Europe/London
BEGIN:DAYLIGHT
TZOFFSETFROM:+0000
TZOFFSETTO:+0100
TZNAME:BST
DTSTART:19700329T010000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0100
TZOFFSETTO:+0000
TZNAME:GMT
DTSTART:19701025T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
CATEGORIES:Microsoft Research Machine Learning and Perception
  Seminars
SUMMARY:Exclusive Pólya Urns and their applications - Chri
 stian Steinruecken
DTSTART;TZID=Europe/London:20111011T140000
DTEND;TZID=Europe/London:20111011T150000
UID:TALK33622AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/33622
DESCRIPTION:The Dirichlet Process (DP) and its variants have m
 any nice mathematical properties which make them p
 opular tools in the machine learning community. Ho
 wever\, deploying DPs can be costly in settings wh
 ere approximate inference over latent states is ne
 eded (e.g. Gibbs sampling seating arrangements in 
 a CRP).\n\nBy dropping some of the "nice" mathemat
 ical properties of the DP\, we can construct a nov
 el kind of stochastic process which offers exact i
 nference\, is fast\, easy to implement\, and can i
 n many cases be used as a drop-in replacement for 
 Chinese Restaurant Processes and Pitman-Yor proces
 ses. I will show how the new process differs from 
 models in the existing literature\, and how it can
  be applied to interesting tasks\, including hiera
 rchical sequence modelling and data compression.
LOCATION:Small public lecture room\, Microsoft Research Ltd
 \, 7 J J Thomson Avenue (Off Madingley Road)\, Cam
 bridge
CONTACT:Microsoft Research Cambridge Talks Admins
END:VEVENT
END:VCALENDAR