Jeudi 31 Mai

Heure:	12:15 - 13:30
Lieu:	Salle B107, bâtiment B, Université de Villetaneuse
Résumé:	Graphes dynamiques - Modélisation et Clustering
Description:	Tabea Rebafka To model recurrent interaction events we propose a new dynamic random graph model : an extension of the stochastic block model in continuous time, where every individual (node) belongs to a latent group and interactions (edges) between two individuals are modelled using inhomogeneous Poisson processes. We develop a semiparametric variational expectation-maximization algorithm to estimate model parameters and to perform a clustering of the nodes. We illustrate the performance of our method on real datasets.