BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Unsupervised Many-to-many Object Matching - Dr Tomoharu Iwata (NTT
 )
DTSTART:20140912T100000Z
DTEND:20140912T110000Z
UID:TALK54243@talks.cam.ac.uk
CONTACT:Zoubin Ghahramani
DESCRIPTION:Object matching is a task for finding correspondences between 
 objects in different domains. Examples of object matching include matching
  an English word with a French word\, and a user identification with a use
 r identification in a different database. Most of the object matching meth
 ods are supervised\; they require similarity measures or paired data. We p
 ropose a probabilistic latent variable model for unsupervised object match
 ing\, which can find many-to-many matching without alignment information. 
 The proposed model assumes that there are an infinite number of latent vec
 tors that are shared by all domains\, and that each object is generated us
 ing one of the latent vectors and a domain-specific linear projection. By 
 inferring a latent vector to be used for generating each object\, objects 
 in different domains are clustered in shared groups\, and thus we can find
  matching between clusters in an unsupervised manner. We present efficient
  inference procedures for the proposed model based on a stochastic EM algo
 rithm. The effectiveness of the proposed model is demonstrated with experi
 ments using synthetic and real data sets.\n
LOCATION:Engineering Department\, CBL Room BE-438
END:VEVENT
END:VCALENDAR