Graph Encoder Ensemble for Simultaneous Vertex Embedding and Community Detection

01/18/2023

∙

by   Cencheng Shen, et al.

∙

0

∙

share

In this paper we propose a novel and computationally efficient method to simultaneously achieve vertex embedding, community detection, and community size determination. By utilizing a normalized one-hot graph encoder and a new rank-based cluster size measure, the proposed graph encoder ensemble algorithm achieves excellent numerical performance throughout a variety of simulations and real data experiments.