BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Graph Kernels for Data Mining - Karsten Borgwardt\, Machine Learni
 ng Group @ CUED
DTSTART:20070919T130000Z
DTEND:20070919T140000Z
UID:TALK8084@talks.cam.ac.uk
CONTACT:Carl Edward Rasmussen
DESCRIPTION:As new graph structured data is constantly being generated\, l
 earning and data mining on graphs have become a challenge in application a
 reas such as molecular biology\, telecommunications\, chemoinformatics\, a
 nd social network analysis. The central algorithmic problem in these areas
 \, measuring similarity of graphs\, has therefore received extensive atten
 tion in the recent past. Unfortunately\, existing approaches are slow\, la
 cking in expressivity\, or hard to parameterize.\n\nGraph kernels have rec
 ently been proposed as a theoretically sound and promising approach to the
  problem of graph comparison. Their attractivity stems from the fact that 
 by defining a kernel on graphs\, a whole family of data mining and machine
  learning algorithms becomes applicable to graphs.\n\nThese kernels on gra
 phs must respect both the information represented by the topology and the 
 node and edge labels of the graphs\, while being efficient to compute. Exi
 sting methods fall woefully short\; they miss out on important\ntopologica
 l information\, are plagued by runtime issues\, and do not scale to large 
 graphs.\n\nIn this talk\, we will present our work on solving these proble
 ms\, and our novel graph kernels and kernel methods for mining large graph
 s and large datasets of graphs.
LOCATION:LR5\, Engineering\, Department of
END:VEVENT
END:VCALENDAR