BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//talks.cam.ac.uk//v3//EN
BEGIN:VTIMEZONE
TZID:Europe/London
BEGIN:DAYLIGHT
TZOFFSETFROM:+0000
TZOFFSETTO:+0100
TZNAME:BST
DTSTART:19700329T010000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0100
TZOFFSETTO:+0000
TZNAME:GMT
DTSTART:19701025T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
CATEGORIES:Machine Learning @ CUED
SUMMARY:Inference as Learning - George Papamakarios (Unive
rsity of Edinburgh)
DTSTART;TZID=Europe/London:20160808T110000
DTEND;TZID=Europe/London:20160808T120000
UID:TALK66984AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/66984
DESCRIPTION:How can we do Bayesian inference if the likelihood
is not available? This situation arises in simula
tor-based models\, where the model can be easily s
imulated but its likelihood is intractable. One ca
n learn to perform inference in such models based
only on simulation data\, by casting inference as
a learning problem. In this talk\, I will describe
a strategy for doing this efficiently using Bayes
ian conditional density estimation\, and compare i
t with established likelihood-free inference techn
iques such as Approximate Bayesian Computation.
LOCATION:CBL Room BE-438
CONTACT:Zoubin Ghahramani
END:VEVENT
END:VCALENDAR