Timezone: »

 
Poster
EvenNet: Ignoring Odd-Hop Neighbors Improves Robustness of Graph Neural Networks
Runlin Lei · Zhen Wang · Yaliang Li · Bolin Ding · Zhewei Wei

@

Graph Neural Networks (GNNs) have received extensive research attention for their promising performance in graph machine learning. Despite their extraordinary predictive accuracy, existing approaches, such as GCN and GPRGNN, are not robust in the face of homophily changes on test graphs, rendering these models vulnerable to graph structural attacks and with limited capacity in generalizing to graphs of varied homophily levels. Although many methods have been proposed to improve the robustness of GNN models, most of these techniques are restricted to the spatial domain and employ complicated defense mechanisms, such as learning new graph structures or calculating edge attentions. In this paper, we study the problem of designing simple and robust GNN models in the spectral domain. We propose EvenNet, a spectral GNN corresponding to an even-polynomial graph filter. Based on our theoretical analysis in both spatial and spectral domains, we demonstrate that EvenNet outperforms full-order models in generalizing across homophilic and heterophilic graphs, implying that ignoring odd-hop neighbors improves the robustness of GNNs. We conduct experiments on both synthetic and real-world datasets to demonstrate the effectiveness of EvenNet. Notably, EvenNet outperforms existing defense models against structural attacks without introducing additional computational costs and maintains competitiveness in traditional node classification tasks on homophilic and heterophilic graphs.

Keywords: graph neural networks · robustness · homophily
[ Paper]  [ Poster]  [ OpenReview] 

Author Information

Runlin Lei (Renmin University of China)
Zhen Wang (Alibaba)

I got my ph.d. from Sun Yat-sen University (a joint program of my school and Microsoft Research Asian). Now, I am working for Alibaba. I had research background on knowledge graph related topics. But now, I am interested in reinforcement learning (RL) and working for a post-doc in RL direction.

Yaliang Li (Alibaba)
Bolin Ding (Alibaba Group)
Zhewei Wei (Renmin University of China)

More from the Same Authors