BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Expectation Propagation on the TrueSkill model - Carl Scheffler (U
 niversity of Cambridge)
DTSTART:20080310T111500Z
DTEND:20080310T121500Z
UID:TALK10973@talks.cam.ac.uk
CONTACT:Carl Scheffler
DESCRIPTION:Over the next couple of weeks we'll look at the TrueSkill syst
 em for ranking players and teams in a competitive game. The main outcome i
 s to understand EP by implementing it.\n\nPlease read at least the first 2
  references below.\n\n*References*\n\n* Ralf Herbrich\, Tom Minka\, Thore 
 Graepel\, ``TrueSkill(TM): A Bayesian Skill Rating System'' ["ps.gz":http:
 //books.nips.cc/papers/files/nips19/NIPS2006_0688.ps.gz] ["pdf":http://boo
 ks.nips.cc/papers/files/nips19/NIPS2006_0688.pdf]\n\n* Pierre Dangauthier\
 , Ralf Herbrich\, Tom Minka\, Thore Graepel\, ``TrueSkill Through Time: Re
 visiting the History of Chess'' ["ps.gz":http://books.nips.cc/papers/files
 /nips20/NIPS2007_0931.ps.gz] ["pdf":http://books.nips.cc/papers/files/nips
 20/NIPS2007_0931.pdf]\n\n* Thomas Minka\, ``A family of algorithms for app
 roximate Bayesian inference''\, Chapters 1 and 3. ["ps.gz":http://research
 .microsoft.com/~minka/papers/ep/minka-thesis.ps.gz] ["pdf":http://research
 .microsoft.com/~minka/papers/ep/defense.pdf]\n
LOCATION:TCM Seminar Room\, Cavendish Laboratory\, Department of Physics
END:VEVENT
END:VCALENDAR