no code implementations • 17 Feb 2023 • Acong Zhang, Jincheng Huang, Ping Li, Kai Zhang
Multiple recent studies show a paradox in graph convolutional networks (GCNs), that is, shallow architectures limit the capability of learning information from high-order neighbors, while deep architectures suffer from over-smoothing or over-squashing.
1 code implementation • 16 Feb 2023 • Jincheng Huang, Lun Du, Xu Chen, Qiang Fu, Shi Han, Dongmei Zhang
Theoretical analyses guarantee the robustness of signals through the mid-pass filter, and we also shed light on the properties of different frequency signals under adversarial attacks.
no code implementations • 23 May 2022 • Jincheng Huang, Ping Li, Rui Huang, Chen Na, Acong Zhang
Alternatively, it is possible to exploit the information about the presence of heterophilous neighbors for feature learning, so a hybrid message passing approach is devised to aggregate homophilious neighbors and diversify heterophilous neighbors based on edge classification.