Seminario: " Model-based clustering for network interaction length data"

Abstract: Network data evolving over time in a continuous fashion are particularly common in a variety of fields where interactions between units are protracted during a certain time interval. For example, data of such type can describe proximity interactions or exchange of phone calls within a collection of individuals, visual contacts between the participants at an event, or movement patterns across different locations in a city. In addition, units tend to create communities that are related to and influence the patterns and the duration of the interactions. In these situations, the main interest is often for how long these entities interact (and conversely do not interact) within the observed time period. This presentation introduces a new stochastic block model that focuses on the analysis of interaction lengths in dynamic networks. The model does not rely on a discretization of the time dimension and may be used to analyze networks that evolve continuously over time. The framework relies on a clustering structure on the nodes, whereby two nodes belonging to the same latent group tend to create interactions and non-interactions of similar lengths. An efficient variational expectation–maximization algorithm is employed to perform inference and the model is showcased in applications including face-to-face interaction data and a bike sharing network.

 

Link TEAMS: clicca qui

 

Referente: Prof. R. Di Mari