GRAND seminar
12:00 noon, Sep 27, Thursday, 2007, by Maurizio Filippone
ST2, 430A

Kernel and Spectral Methods for Clustering

Abstract

Clustering algorithms are a useful tool to explore data structures and have been employed in many disciplines. The focus of this talk is the partitioning clustering problem with a special interest in two recent approaches: kernel and spectral methods. The aim is to present a brief survey of kernel and spectral clustering methods, two approaches able to produce nonlinear separating hypersurfaces between clusters. The relationships between these seemingly different paradigms is discussed since it has been recently shown that they have the same mathematical foundation.

Biography

Maurizio Filippone received a Laurea degree in Physics in 2004 from the University of Genova and now he is pursuing a PhD in Computer Science at University of Genova. His research interests are focused on Machine Learning techniques and their applications to Bioinformatics and time series analysis and forecasting.




Department of Computer Science
Volgenau School of Information Technology and Engineering
George Mason University