BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:CBL Alumni Talk: Accurate Gaussian Processes and how they can help
  Deep Learning - Mark van der Wilk\, Imperial College London
DTSTART:20210430T150000Z
DTEND:20210430T160000Z
UID:TALK158821@talks.cam.ac.uk
CONTACT:Elre Oldewage
DESCRIPTION:In my opinion\, model selection is the most appealing capabili
 ty of Bayesian inference\, which has the most to offer in deep learning. H
 owever\, performing Bayesian model selection requires accurate approximate
  inference. In the first part of the talk\, I will discuss accurate infere
 nce in the fundamental building block of deep neural networks: a single la
 yer. Specifically\, I will focus on the Gaussian process (GP) representati
 on of neural network layers\, and present some recent work on inducing poi
 nt and conjugate gradient approximations\, while paying close attention to
  the question of what we should expect from methods that we consider "good
 " or even "exact". In the second part of the talk\, I will discuss how the
 se techniques can be of use in model selection in deep learning\, with exa
 mples on learning invariances. I will close off with some thoughts on how 
 these ideas may develop in the future.
LOCATION:https://eng-cam.zoom.us/j/82969702755?pwd=L0dIVnlwSHJHV2NGbUQ1cmx
 pYjIyUT09
END:VEVENT
END:VCALENDAR