MLGAL: Multi-level Label Graph Adaptive Learning for node clustering in the attributed graph

Jiajun Yu, Adele Lu Jia

Abstract: Highlights•We propose a novel method (MLGAL) for node clustering in the Attributed Graph.•We design an effective graph filter to maximize the smoothness of the node features.•We propose the binary-class pseudo label for graph adaptive learning.•Various threshold update functions are utilized in graph adaptive learning.•MLGAL gets better performance than previous start-of-the-art approaches.