A Game-Theoretic Model for Grouping
Professor Andrea Torsello ( Ca' Foscari University of Venice)
NICTA SML SEMINARDATE: 2007-12-20
TIME: 10:00:00 - 11:00:00
LOCATION: RSISE Seminar Room, ground floor, building 115, cnr. North and Daley Roads, ANU
CONTACT: JavaScript must be enabled to display this email address.
ABSTRACT:
We present a novel framework for grouping based on the concept of competition for class membership. According to this perspective, the focus shifts from optima of objective functions to equilibria of games, allowing us to deal with asymmetric and/or negative affinities. The approach generalizes the the graph-theoretic concept of a dominant-set, which has been shown to provide a robust framework for clustering. Finally, we show how the some properties of the approach can be used to enumerate possibly overlapping groups and cliques.
BIO:
Andrea Torsello received his PhD in computer science at the University of York, UK, and is currently an assistant professor at Ca' Foscari University of Venice, Italy. His research interests are in the areas of computer vision and pattern recognition, in particular, the interplay between stochastic and structural approaches as well as game-theoretic models. Recently, he co-edited a special issue of Pattern Recognition on "Similarity-based pattern recognition". Dr Torsello has published around 40 technical papers in refereed journals and conference proceedings and has been in the program committees of various international conferences and workshops.
