BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Variational inference for scalable Gaussian process approximations
  - Alexander Matthews (University of Cambridge)
DTSTART:20160602T103000Z
DTEND:20160602T113000Z
UID:TALK66446@talks.cam.ac.uk
CONTACT:Louise Segar
DESCRIPTION:Gaussian processes are heavily used as nonparametric priors on
  functions. There are two challenges that are often relevant in this area:
  dealing with non-Gaussian likelihoods and scaling inference.\n\nIn this t
 alk we discuss recent progress in using variational inference to meet thes
 e challenges. In the first part of the talk we resolve some theoretical is
 sues around variational inference in infinite dimensional models. In the s
 econd part we give a variety of practical examples of the use of these app
 roximations and share insights. Finally we will discuss GPflow a software 
 library that implements these ideas using TensorFlow. 
LOCATION:Engineering Department\, CBL Room BE-438
END:VEVENT
END:VCALENDAR