Entropy-based random models for hypergraphs

Network science has traditionally focused on pairwise relationships while disregarding many-body interactions. Hypergraphs are promising mathematical objects for the description of the latter ones. Here, we propose null models to analyse hypergraphs that generalise the classical Erdös-Rényi and Conf...

Ful tanımlama

Detaylı Bibliyografya
Asıl Yazarlar: Saracco, F, Petri, G, Lambiotte, R, Squartini, T
Materyal Türü: Internet publication
Dil:English
Baskı/Yayın Bilgisi: 2022
Diğer Bilgiler
Özet:Network science has traditionally focused on pairwise relationships while disregarding many-body interactions. Hypergraphs are promising mathematical objects for the description of the latter ones. Here, we propose null models to analyse hypergraphs that generalise the classical Erdös-Rényi and Configuration Model by randomising incidence matrices in a constrained fashion. After discussing them, we extend the definition of several network quantities to hypergraphs, derive their expected values and compare them with empirical ones, to detect significant deviations from random behaviours.