BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Random projection ensemble classification - Tim Cannings (Universi
 ty of Cambridge)
DTSTART:20151118T160000Z
DTEND:20151118T170000Z
UID:TALK61367@talks.cam.ac.uk
CONTACT:Adam Kashlak
DESCRIPTION:We introduce a very general method for high-dimensional classi
 fication\, based on careful combination of the results of applying an arbi
 trary base classifier to random projections of the feature vectors into a 
 lower-dimensional space. In one special case presented here\, the random p
 rojections are divided into non-overlapping blocks\, and within each block
  we select the projection yielding the smallest estimate of the test error
 . Our random projection ensemble classifier then aggregates the results of
  applying the base classifier on the selected projections\, with a data-dr
 iven voting threshold to determine the final assignment.  Our theoretical 
 results elucidate the effect on performance of increasing the number of pr
 ojections.  Moreover\, under a boundary condition implied by the sufficien
 t dimension reduction assumption\, we control the test excess risk of the 
 random projection ensemble classifier.  A simulation comparison with sever
 al other popular high-dimensional classifiers reveals its excellent finite
 -sample performance.  This is joint work with Richard Samworth.
LOCATION:MR14\, Centre for Mathematical Sciences
END:VEVENT
END:VCALENDAR