Immunizing over-smoothing: Dual-channel graph convolutional network

doi:10.1016/j.eswa.2025.128976

Immunizing over-smoothing: Dual-channel graph convolutional network

2025-07-16

0

PRE

AI

DOI：10.1016/j.eswa.2025.128976

Original

Original request for help

Save

Abstract

En 中文

• Establish a new framework to alleviate the over-smoothing issue in deep GCNs. • Two subnetworks complement each other to mitigate the structural degradation. • Design a collaborative loss to achieve the effective integration of features.

Keywords:

over-smoothing

graph convolutional networks

subnetworks

structural degradation

collaborative loss

Immunizing over-smoothing: Dual-channel graph convolutional network

Abstract

Journal

Organization