Achieving the Kesten-Stigum bound in the non-uniform hypergraph stochastic block model
Researchers have developed a new spectral algorithm for community detection in non-uniform hypergraphs, which can capture complex, multi-view interactions. This algorithm achieves a Kesten-Stigum-type bound for weak recovery, confirming a conjecture for models with two blocks. The method utilizes an optimally weighted non-backtracking operator and a novel Ihara-Bass formula to efficiently cluster data with heterogeneous higher-order interactions. AI
IMPACT Introduces a novel spectral algorithm for clustering in non-uniform hypergraphs, potentially improving analysis of complex relational data.