BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Energy-Based Models - Stratis Markou and James Allingham (Universi
 ty of Cambridge)
DTSTART:20211201T110000Z
DTEND:20211201T123000Z
UID:TALK165922@talks.cam.ac.uk
CONTACT:Elre Oldewage
DESCRIPTION:While the log-likelihood of the data is analytic in many famil
 iar machine learning models\, there exist certain models\, called Energy B
 ased Models (EBMs)\, for which the log-likelihood is intractable. Lifting 
 the restriction of a tractable log-likelihood opens the door to great flex
 ibility in model architecture and design\, but naturally complicates train
 ing. In this presentation\, we provide an overview of both established as 
 well as recently developed methods for training EBMs. \n\nReading list\nHo
 w to train your energy based model (https://arxiv.org/pdf/2101.03288.pdf)\
 n\nOur reading groups are livestreamed via Zoom and recorded for our Youtu
 be channel. The Zoom details are distributed via our weekly mailing list.
LOCATION: Cambridge University Engineering Department \,LR3A
END:VEVENT
END:VCALENDAR