New paper!
How can we detect the presence of communities in networks with higher-order interactions? For instance, by maximizing hypermodularity! Also, this formulation will allow you to leverage tensor spectral methods to do it. Additionally, the paper also argues that the "overfitting" of modularity methods is actually just people applying them where they are not supposed to be used. And, as a byproduct, there is an explanation of why higher-order SVD works so well in classification tasks in machine learning. Oh, the code is available to use in your own projects (link in the first comment). And moreover, the code includes an efficient data structure for higher-order networks that is independent from the community detection method and that you can also use in your own work. đ
https://journals.aps.org/prresearch/abstract/10.1103/58dr-wktc
#networks #complexity #physics #maths #CompSci #graphs #higherorder #hypergraphs #community #detection #algorithm #communitystructure #modularity #hypermodularity