Timezone: »

Scattering GCN: Overcoming Oversmoothness in Graph Convolutional Networks
Yimeng Min · Frederik Wenkel · Guy Wolf

Thu Dec 10 09:00 AM -- 11:00 AM (PST) @ Poster Session 5 #1426

Graph convolutional networks (GCNs) have shown promising results in processing graph data by extracting structure-aware features. This gave rise to extensive work in geometric deep learning, focusing on designing network architectures that ensure neuron activations conform to regularity patterns within the input graph. However, in most cases the graph structure is only accounted for by considering the similarity of activations between adjacent nodes, which limits the capabilities of such methods to discriminate between nodes in a graph. Here, we propose to augment conventional GCNs with geometric scattering transforms and residual convolutions. The former enables band-pass filtering of graph signals, thus alleviating the so-called oversmoothing often encountered in GCNs, while the latter is introduced to clear the resulting features of high-frequency noise. We establish the advantages of the presented Scattering GCN with both theoretical results establishing the complementary benefits of scattering and GCN features, as well as experimental results showing the benefits of our method compared to leading graph neural networks for semi-supervised node classification, including the recently proposed GAT network that typically alleviates oversmoothing using graph attention mechanisms.

Author Information

Yimeng Min (Mila)
Frederik Wenkel (Mila, Université de Montréal)

Frederik Wenkel is a PhD candidate in Applied Mathematics at Université de Montréal and Mila (the Quebec AI institute) working on Geometric Deep Learning. In particular, he is interested in Graph Neural Networks and their applications in domains such as Social Networks, Biochemistry and Finance. He holds a bachelor’s and master’s degree in Mathematics at Technical University of Munich.

Guy Wolf (Université de Motréal; Mila)

More from the Same Authors