Flexible Models for Complex Networks
The first generation of models for complex networks consisted of random graphs with power law degree distributions and the small world property (clustering and low diameter). However, as more examples are observed in the applications context, objections have been raised concerning the lack of flexibility of such random graph models.
In this talk we will present flexible models for complex networks, including models that generate graphs with negative assortativity (this is a generalization of the Erdos-Gallai and Havel-Hakimi method), geometric dot product graphs for a wide range of densities and generating distributions (this is a vast generalization of classical Erdos-Renyi graphs), and graphs arising from network formation games.
Milena Mihail, Georgia Institute of Technology