Convolutional Neural Networks on Graphs with Chebyshev Approximation, Revisited

arXiv:2202.03580v5 Announce Type: replace
Abstract: Designing spectral convolutional networks is a challenging problem in graph learning. ChebNet, one of the early attempts, approximates the spectral graph convolutions using Chebyshev polynomials. GCN simplifies ChebNet by utilizing only the first two Chebyshev polynomials while still outperforming it on real-world datasets. GPR-GNN and BernNet demonstrate that the Monomial and Bernstein bases also outperform the Chebyshev basis in terms of learning the spectral graph convolutions. Such conclusions are counter-intuitive in the field of approximation …

abstract approximation arxiv convolutional neural networks cs.ai cs.lg datasets designing gnn graph graph learning graphs networks neural networks type world