Spectral Graph Convolutional Networks With Lifting-based Adaptive Graph Wavelets

08/03/2021
by   Mingxing Xu, et al.
6

Spectral graph convolutional networks (SGCNs) have been attracting increasing attention in graph representation learning partly due to their interpretability through the prism of the established graph signal processing framework. However, existing SGCNs are limited in implementing graph convolutions with rigid transforms that could not adapt to signals residing on graphs and tasks at hand. In this paper, we propose a novel class of spectral graph convolutional networks that implement graph convolutions with adaptive graph wavelets. Specifically, the adaptive graph wavelets are learned with neural network-parameterized lifting structures, where structure-aware attention-based lifting operations are developed to jointly consider graph structures and node features. We propose to lift based on diffusion wavelets to alleviate the structural information loss induced by partitioning non-bipartite graphs. By design, the locality and sparsity of the resulting wavelet transform as well as the scalability of the lifting structure for large and varying-size graphs are guaranteed. We further derive a soft-thresholding filtering operation by learning sparse graph representations in terms of the learned wavelets, which improves the scalability and interpretablity, and yield a localized, efficient and scalable spectral graph convolution. To ensure that the learned graph representations are invariant to node permutations, a layer is employed at the input of the networks to reorder the nodes according to their local topology information. We evaluate the proposed networks in both node-level and graph-level representation learning tasks on benchmark citation and bioinformatics graph datasets. Extensive experiments demonstrate the superiority of the proposed networks over existing SGCNs in terms of accuracy, efficiency and scalability.

READ FULL TEXT

page 1

page 11

research
10/12/2020

Locality Preserving Dense Graph Convolutional Networks with Graph Context-Aware Node Representations

Graph convolutional networks (GCNs) have been widely used for representa...
research
04/26/2019

Graph Optimized Convolutional Networks

Graph Convolutional Networks (GCNs) have been widely studied for graph d...
research
09/09/2016

Semi-Supervised Classification with Graph Convolutional Networks

We present a scalable approach for semi-supervised learning on graph-str...
research
06/27/2019

Adversarial Representation Learning on Large-Scale Bipartite Graphs

Graph representation on large-scale bipartite graphs is central for a va...
research
07/30/2020

grid2vec: Learning Efficient Visual Representations via Flexible Grid-Graphs

We propose grid2vec, a novel approach for image representation learning ...
research
12/12/2019

Tracing the Propagation Path: A Flow Perspective of Representation Learning on Graphs

Graph Convolutional Networks (GCNs) have gained significant developments...
research
06/19/2020

Localized Spectral Graph Filter Frames: A Unifying Framework, Survey of Design Considerations, and Numerical Comparison

Representing data residing on a graph as a linear combination of buildin...

Please sign up or login with your details

Forgot password? Click here to reset