Tag
This paper proposes SCISE, a scalable unsupervised graph clustering framework that uses community-aware sampling and structural entropy to overcome structural isolation in mini-batch training, achieving state-of-the-art results on benchmark datasets.
Introduces SNMPBB, a nonmonotone gradient-based algorithm for symmetric nonnegative matrix factorization that achieves significant speedups over existing methods, with extensions to graph clustering and low-rank approximations.