Clustering dynamic random graphs

Gio, 19/05/2022 - 12:30 / 13:30

405, Viale Romania

Speaker: Catherine Matias , CNRS, Sorbonne Université & Université Paris Cité

Abstract

Interactions between individuals are nowadays often collected and available in a dynamic / temporal setting. I’ll start by discussing the different types of dynamic interaction data. Then I’ll discuss clustering techniques that help summarize the individuals into groups with similar interaction behaviours. I’ll focus mostly on the Stochastic Blockmodel and its versions for temporal interactions datasets. I’ll illustrate these concepts and their use on two different datasets: an ecological network of interactions between ants and a dataset of cycles journeys between cycle hiring stations in London. 

Further info available at: