The problem of detecting community structure in complex networks received quite a bit of attention in recent years, probably since it is not only theoretically important, but also highly relevant to practical applications. Therefore, a wide range of tools have been applied to this domain, including both more traditional machine learning methods, and various types of evolutionary algorithms. In this tutorial we will not only dive into some of the fascinating methods employed in this highly active field, but also compare them in various ways, using several different yardsticks. On top of that, we will attempting to answer the question: have evolutionary methods proven their worth in this complex domain, or is it currently better to rely on standard clustering methods?


Ami Hauptman Ami Hauptman holds a PhD from Ben-Gurion University. His thesis focused on applying evolutionary methods to complex games, and has won several Humies awards, including two gold awards. In the past 6 years he has been employed at some top-notch Israeli companies, mainly in applying learning methods to difficult problems, including cyber attack detection, social network analysis and classification of images. Recently, he has returned to academia to carry on his research and share his knowledge with others.

Evolutionary computation methods e-session