The reviewed record of science sign in
Pith

arxiv: 2202.13526 · v2 · pith:4TGBLIDR · submitted 2022-02-28 · cs.LG · eess.SP

Sparse Graph Learning with Spectrum Prior for Deep Graph Convolutional Networks

Reviewed by Pith T0 review T1 audit T2 compute T3 formal T4 kernel pith:4TGBLIDRrecord.jsonopen to challenge →

classification cs.LG eess.SP
keywords graphspectrumpriordatalearningoutputover-smoothingsparse
0
0 comments X
read the original abstract

A graph convolutional network (GCN) employs a graph filtering kernel tailored for data with irregular structures. However, simply stacking more GCN layers does not improve performance; instead, the output converges to an uninformative low-dimensional subspace, where the convergence rate is characterized by the graph spectrum -- this is the known over-smoothing problem in GCN. In this paper, we propose a sparse graph learning algorithm incorporating a new spectrum prior to compute a graph topology that circumvents over-smoothing while preserving pairwise correlations inherent in data. Specifically, based on a spectral analysis of multilayer GCN output, we derive a spectrum prior for the graph Laplacian matrix $\mathbf{L}$ to robustify the model expressiveness against over-smoothing. Then, we formulate a sparse graph learning problem with the spectrum prior, solved efficiently via block coordinate descent (BCD). Moreover, we optimize the weight parameter trading off the fidelity term with the spectrum prior, based on data smoothness on the original graph learned without spectrum manipulation. The output $\mathbf{L}$ is then normalized for supervised GCN training. Experiments show that our proposal produced deeper GCNs and higher prediction accuracy for regression and classification tasks compared to competing schemes.

This paper has not been read by Pith yet.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.