BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Bayesian Generative Adversarial Networks - Professor Andrew Wilson
 \, Cornell University
DTSTART:20171213T110000Z
DTEND:20171213T120000Z
UID:TALK94141@talks.cam.ac.uk
CONTACT:Pat Wilson
DESCRIPTION:Through an adversarial game\, generative adversarial networks 
 (GANs) can implicitly learn rich distributions over images\, audio\, and d
 ata which are hard to model with an explicit likelihood.  I will present a
  practical Bayesian formulation for unsupervised and semi-supervised learn
 ing with GANs.  Within this framework\, we use a stochastic gradient Hamil
 tonian Monte Carlo for marginalizing parameters.  The resulting approach c
 an automatically discover complementary and interpretable generative hypot
 heses for collections of images\, without ad hoc interventions.  Moreover\
 , by exploring an expressive posterior over these hypotheses\, we show tha
 t it is possible to achieve state-of-the-art quantitative results on image
  classification benchmarks\, even with less than 1% of the labelled traini
 ng data.  
LOCATION:CBL Seminar Room
END:VEVENT
END:VCALENDAR