BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Bayesian Ranking - Ralf Herbrich\, Microsoft Research Cambridge
DTSTART:20070315T160000Z
DTEND:20070315T180000Z
UID:TALK6598@talks.cam.ac.uk
CONTACT:Zoubin Ghahramani
DESCRIPTION:In this talk I will present a Bayesian approach to ranking a s
 et of objects based on the possibly partial or noisy rankings of small sub
 sets of objects. Rankings are represented by assigning a latent real-value
 d variable (skill\, urgency\, value) to each object and sorting the object
 s according to the magnitude of the latent variables. The system maintains
  a Gaussian belief about the value of each object in terms of mean and var
 iance. I will discuss approximate message passing in factor graphs as the 
 computational technique to address the problem of inference.\n\nAfter pres
 enting theoretical and algorithmic aspects of the system\, I will outline 
 two applications:\n\n* TrueSkill(TM) - Ranking of players: The system is u
 sed to provide matchmaking and leaderboard functionality based on the esti
 mated skills of players. An implementation of the system is currently at t
 he heart of ranking and matchmaking in the online gaming service Xbox Live
 \, used by 1 million players playing approximately 500\,000 ranked matches
  every 24 hours.\n\n* Liberty - Ranking of potential moves in Computer Go:
  The system is used to learn the values of local patterns based on the mov
 es played in a given position. The "winner" is determined by observing whi
 ch one of the legal moves in a given position has been played by an expert
  player. The result is a probability distribution over moves for a given p
 osition. It can serve\, for example\, as fast stand-alone Go engine of res
 pectable playing strength. The current system plays at 10-15 kyu and corre
 ctly predicts expert moves in 34% of the cases.\n
LOCATION:LR4\, Engineering\, Department of
END:VEVENT
END:VCALENDAR