BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Vector Gaussian Processes - Ryan Adams\, Inference Group
DTSTART:20061115T140000Z
DTEND:20061115T150000Z
UID:TALK5715@talks.cam.ac.uk
CONTACT:Ryan Prescott Adams
DESCRIPTION:Gaussian process regression is often compared to the use of ne
 ural networks for function approxmation.  One significant difference\, how
 ever\, is that neural networks straightforwardly extend to multiple output
 s.  Constructing nontrivial (anti)correlations in Gaussian process predict
 ive distributions turns out to be somewhat difficult.  I will talk about o
 ne way to approach this problem.
LOCATION:Ryle Seminar Room\, Cavendish Laboratory
END:VEVENT
END:VCALENDAR