Aad van der Vaart: Statistical estimation in preferential attachment models

(joint with Fengnan Gao, Fudan University)

Preferential attachment models are dynamic network models in which new nodes are attached to nodes in an existing graph that are randomly chosen by a probability that is a function of the degree of the nodes. We consider estimating the attachment function, which governs the probabilities, from the observed graph, where we consider both parametric models and nonparametric models.

We show consistency, and in some cases asymptotic normality, of the estimators using the theory of supercritical branching processes and urn models. We shall end the talk with some open problems.