Talk by Giona Casiraghi, ETH Zurich

From biased urns to temporal networks: modeling multi-edge graphs with the generalized hypergeometric ensemble

Thursday 14.03.24  Time: 12.15 - 13.45  Room: Y27H12

Abstract:

Empirical networks are characterized by multi-edges--repeated interactions between the same node pairs. Ignoring this characteristic in their analysis may yield erroneous conclusions. Our research demonstrates the effectiveness of biased urn models for capturing the features of complex networks, particularly in the context of multi-edge formation and evolution. By means of the Generalised Hypergeometric Ensemble (gHypE), we provide a robust analytical framework for statistical inference on multi-edged networks. This framework allows to accurately capture empirical network properties and significantly enhances our understanding of networks through adaptable and computationally efficient modeling techniques.