BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Markov Random Fields for Classification in High Dimensional Spaces
  with Application to fMRI Analysis - Dr Avishy Carmi\, CUED
DTSTART:20090121T141500Z
DTEND:20090121T151500Z
UID:TALK16600@talks.cam.ac.uk
CONTACT:Rachel Fogg
DESCRIPTION:In this talk we present a new classification algorithm for hig
 h dimensional problems. The algorithm uses a Markov random field for model
 ing meaningful interactions within the training data set.\nThe model param
 eters are efficiently estimated using the Kalman filter algorithm and adap
 ted to fit the test data using a recursive matrix formulation of the exten
 ded Baum-Welch algorithm. A spatially likelihood test procedure is then us
 ed for classifying the data. The performance of the new algorithm is demon
 strated in fMRI classification.\n
LOCATION:LR5\, Engineering\, Department of
END:VEVENT
END:VCALENDAR