BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Efficient Machine Learning with High Order and Combinatorial Struc
 tures - Danny Tarlow
DTSTART:20120405T083000Z
DTEND:20120405T093000Z
UID:TALK37296@talks.cam.ac.uk
CONTACT:Microsoft Research Cambridge Talks Admins
DESCRIPTION:Two challenges facing machine learning researchers are (a) how
  to build models\nthat more accurately reflect prior beliefs and constrain
 ts relevant to the\ndomain being modeled\, and (b) how to model more compl
 exly structured data.  In\nboth of these cases\, there are tradeoffs betwe
 en the expressibility of the model\nand the computational efficiency of le
 arning and inference procedures. In this\ntalk\, I will discuss several re
 cent approaches towards building more expressive\nand more efficient model
 s of structured domains.  The focus will be on learning\nand inference pro
 cedures for models that have non-local (high order) and/or\ncombinatorial 
 structure.\n
LOCATION:Large lecture theatre\, Microsoft Research Ltd\, 7 J J Thomson Av
 enue (Off Madingley Road)\, Cambridge
END:VEVENT
END:VCALENDAR