Northeastern University
Sign Up
Free Event

Title: The Trouble with Community Detection

Speaker: Dr. Aaron Clauset, Assistant Professor of Computer Science at the University of Colorado at Boulder

Location: 177 Huntington Avenue, 11th Floor, Boston, Massachusetts 02115

Abstract

A common task in network analysis is to seek a coarse-graining of the network into modules or communities, which describe the large-scale architecture of the network.  For instance, we might want to find social groups within a network of friendships, functional modules among gene regulatory interactions, or compartments within food webs.  However, different algorithms will return different communities for the same network, and this presents a conundrum for scientific interpretation: which set of communities are the real ones?

In this talk, Dr. Clauset will show how using node attributes or "metadata" can solve this problem, by guiding the community detection process toward useful outcomes.  The resulting algorithm, which is a generalization of the powerful stochastic block model, is more accurate than any algorithm that uses only network structure or node metadata alone, and can automatically learn the underlying correlation between metadata and structure, if one exists.  To illustrate these features, Clauset will show results for applying the method both to synthetic networks with known structure and to real-world networks with unknown structure.  He will close with a few general comments about the recently proved 'No Free Lunch' theorem in community detection, and the utility of community detection methods in scientific applications.

This is joint work with Mark Newman.

About the Speaker 

Aaron Clauset's group research activities are broad and multidisciplinary, and are active participants in the network science, complex systems, computational biology, and computational social science communities.  His work generally focuses on understanding the mechanisms by which large-scale patterns emerge from the collective actions of heterogeneous individuals and on developing novel techniques for inferring such patterns and mechanisms from rich data sources.  Much of this work is methdological in nature, and he actively develops novel statistical and computational methods for automatically analyzing and modeling complex phenomena in biological, social and technological systems.  All of these efforts draw heavily on data analysis, machine learning, statistics, probability, algorithms and graph theory.  He is particularly interested in interactions between theory and data, and the development of rigorous methods for the study of complex systems.

Three areas of current work are the large-scale organization of complex networks, with particular emphasis on their modular or hierarchical structure; the mechanisms that shape the macroevolution of biological species across large spatial and temporal scales; and the political and physical processes that shape the large-scale dynamics of violent human conflicts, such as modern terrorism and warfare.

**** Please bring your Northeastern University ID (or other photo ID) when entering the building. **** 

Event Details

0 people are interested in this event

User Activity

No recent activity