Federated graph-level clustering network with adaptive knowledge compensation

Renda Han

doi:10.1016/j.neunet.2025.108212

Federated graph-level clustering network with adaptive knowledge compensation

2025-10-22

0

PRE

AI

DOI：10.1016/j.neunet.2025.108212

Original

Original request for help

Save

Abstract

En 中文

• Local Knowledge Extraction: Clients extract reliable knowledge from multi-source data. • Local Refinement and Prototype Correction: Refined knowledge is enhanced using global prototype correction. • Global Prototype Alignment: The server builds adaptive affinities to group clients into semantic communities.

Federated graph-level clustering network with adaptive knowledge compensation

Abstract

Journal

Organization