BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Unsupervised Representation Learning - Amar Shah (University of Ca
 mbridge)
DTSTART:20140612T140000Z
DTEND:20140612T153000Z
UID:TALK53033@talks.cam.ac.uk
CONTACT:Konstantina Palla
DESCRIPTION:Being able to learn 'good' representations of data is\, arguab
 ly\, 90% \nof the hard work for machine learning tasks. We currently have 
  an \nabundance of unlabelled data\, with more being created every day. \n
 It is therefore imperative that we can design and train representation \nl
 earning algorithms in an unsupervised setting.\n \nIn this tutorial style 
 talk\, we explore probabilistic and non-\nprobabilistic approaches includi
 ng Principal Component Analysis\,\nRestricted Boltzman Machines and Autoen
 coders. We will also\ndiscuss the benefits of deep architectures\, and how
  to go about \ntraining them.
LOCATION:Engineering Department\, CBL Room 438
END:VEVENT
END:VCALENDAR