BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Expectation Propagation in Sparse Linear Models with Spike and Sla
 b Priors - Dr José Miguel Hernández Lobato (Univ. Aut. Madrid)
DTSTART:20110304T113000Z
DTEND:20110304T123000Z
UID:TALK30017@talks.cam.ac.uk
CONTACT:Zoubin Ghahramani
DESCRIPTION:Sparse linear models assume that the data have been generated 
 by a linear model whose\ncoefficient vector is sparse: a small number of c
 oefficients take values that\nare significantly different from zero\, whil
 e the remaining coefficients are exactly zero.\nThis configuration is espe
 cially useful for addressing learning problems with a small number of trai
 ning\ninstances and a high-dimensional feature space. In a Bayesian approa
 ch\, sparsity can be favored by using\nspecific priors such as the spike a
 nd slab distribution. In this talk\, different sparse\nlinear models with 
 spike and slab priors will be analyzed\, using expectation propagation for
  fast approximate inference.
LOCATION:Engineering Department\, CBL Room 438
END:VEVENT
END:VCALENDAR