BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:If Bayesian inference is rational\, what is approximate inference?
  - Sebastian Farquhar
DTSTART:20220426T140000Z
DTEND:20220426T150000Z
UID:TALK173126@talks.cam.ac.uk
CONTACT:96082
DESCRIPTION:There are some good reasons to like Bayesian inference\, but t
 hese are awkwardly mostly also reasons to dislike *approximate* Bayesian i
 nference. Nevertheless\, we need approximations. So are there any ways to 
 understand what would make one approximate inference scheme better than an
 other? I argue that this is very hard. Sequential learning problems like a
 ctive and continual learning could be seen as evaluations of approximate i
 nference\, but doing this carefully is more subtle than it at first seems.
LOCATION:SS03\, Computer Lab
END:VEVENT
END:VCALENDAR