BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Efficient Sampling with Kernel Herding - Yutian Chen (University o
 f California at Irvine) - talk given by videolink
DTSTART:20120912T150000Z
DTEND:20120912T160000Z
UID:TALK39838@talks.cam.ac.uk
CONTACT:Zoubin Ghahramani
DESCRIPTION:The herding algorithm was proposed as a deterministic\ndynamic
  system that integrates learning and inference for discrete\nMarkov Random
  Fields. In this talk\, we take a dual view of herding and\nextend it to c
 ontinuous spaces by using the kernel trick. The\nresulting “kernel herdi
 ng” is an infinite memory deterministic\nalgorithm that approximates a P
 DF with a collection of samples. We\nshow that kernel herding decreases th
 e error of expectations of\nfunctions in the Hilbert space much faster tha
 n the usual iid random\nsamples. If time permits\, I’ll also talk about 
 the recent development\nand applications of the herding algorithm.
LOCATION:Engineering Department\, CBL Room BE-438
END:VEVENT
END:VCALENDAR