Seminar

A fundamental problem in the study of real-world networks is how to discover topological properties of a network given limited time and resources. For networks which are not too large, a random walker with stationary transition probabilities is usually sufficient. Yet, these methods will not scale up to large graphs, and especially not to modular and directed graphs. I will discuss methods of network discovery in large networks and also in model networks that share features with adversarial networks such as crime or terrorist networks. I will also discuss the use of multiple walkers as a possible analytical tool for studying walks on directed networks.