BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Multiscale Methods for the Analysis of Dynamic Graphs - Maggioni\,
  M (Duke)
DTSTART:20100625T084500Z
DTEND:20100625T093000Z
UID:TALK25333@talks.cam.ac.uk
CONTACT:Mustapha Amrani
DESCRIPTION:Dynamic graphs arise in a variety of real-world situations: fr
 om social networks\, to engineered\nphysical networks\, to graphs associat
 ed with data sets (e.g. financial transactions) that vary in\ntime. The ch
 allenges are the need to develop robust tools and metrics for comparing gr
 aphs at different\ntimes\, in order to model statistical significant chang
 es\, and capture anomalies: in real-world\nsituation a graph/network will 
 vary stochastically in time with vertex/edge additions/deletions\,\nand cl
 assical tools such as graph isomorphism are not robust enough to handle su
 ch changes. We\nuse multiscale decompositions of graph and random walks at
  multiple scales to introduce metrics\nof change (in time) of a graph\, th
 at allow use to capture changes of different magnitude at different\nscale
 s and locations on the graph. We apply these techniques to synthetic graph
 s as well as real\nworld data sets\, and discuss strengths and weaknesses 
 of this approach.
LOCATION:Seminar Room 1\, Newton Institute
END:VEVENT
END:VCALENDAR