BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Bayesian regression and classification with multivariate sparsifyi
 ng priors - Prof. Tom Heskes (Radboud University Nijmegen)
DTSTART:20110621T133000Z
DTEND:20110621T143000Z
UID:TALK31840@talks.cam.ac.uk
CONTACT:Zoubin Ghahramani
DESCRIPTION:Many regression and classification problems in neuroimaging an
 d bioinformatics belong to the class "large p\, small n": many variables\,
  just a few data points. Popular methods for handling such problems includ
 e L1-regularization and spike-and-slab variable selection. These methods a
 re univariate when it comes to determine which variables are selected. In 
 this talk I will present multivariate extensions that allow  for the incor
 poration of (spatio-temporal) constraints and lead to smooth importance ma
 ps. I will discuss how to arrive at efficient algorithms for (approximate)
  inference and will illustrate the methods on fMRI analysis and EEG source
  localization.
LOCATION:Engineering Department\, CBL Room 438
END:VEVENT
END:VCALENDAR