BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Stable distribution and data sketching - Ioana Cosma\, Department 
 of Statistics\, University of Oxford
DTSTART:20080206T140000Z
DTEND:20080206T150000Z
UID:TALK9678@talks.cam.ac.uk
CONTACT:Zoubin Ghahramani
DESCRIPTION:We introduce the stable law and its properties\, and present a
 n interesting application in dimension reduction over data streams known a
 s _data sketching_.  Of particular interest is the situation where the dat
 a stream is prohibitively large that it cannot be stored in main computer 
 memory or on disk for subsequent access\, but instead must be processed on
  the fly.  The method of data sketching constructs a lower dimensional rep
 resentation such that l_{alpha} norms and quasi-norms of the data stream\,
  for 0 < alpha <= 2\, can be accurately recovered with significantly reduc
 ed computational cost.  In the extreme case as alpha tends to 0\, the norm
  converges to the Hamming distance\, giving the cardinality of the data st
 ream\, while for alpha >= 1\, the norm of the difference between two strea
 ms is a meaningful measure of dissimilarity.
LOCATION:Engineering Department\, CBL Room 438
END:VEVENT
END:VCALENDAR