DFNets: Spectral CNNs for Graphs with Feedback-Looped Filters

NeurIPS 2019  ·  Asiri Wijesinghe, Qing Wang ·

We propose a novel spectral convolutional neural network (CNN) model on graph structured data, namely Distributed Feedback-Looped Networks (DFNets). This model is incorporated with a robust class of spectral graph filters, called feedback-looped filters, to provide better localization on vertices, while still attaining fast convergence and linear memory requirements. Theoretically, feedback-looped filters can guarantee convergence w.r.t. a specified error bound, and be applied universally to any graph without knowing its structure. Furthermore, the propagation rule of this model can diversify features from the preceding layers to produce strong gradient flows. We have evaluated our model using two benchmark tasks: semi-supervised document classification on citation networks and semi-supervised entity classification on a knowledge graph. The experimental results show that our model considerably outperforms the state-of-the-art methods in both benchmark tasks over all datasets.

PDF Abstract NeurIPS 2019 PDF NeurIPS 2019 Abstract

Results from the Paper

Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Node Classification Citeseer DFNet-ATT Accuracy 74.7 ± 0.4 # 26
Node Classification Cora DFNet-ATT Accuracy 86% ± 0.4% # 20
Node Classification NELL DFNet-ATT Accuracy 68.8 ± 0.3 # 1
Node Classification Pubmed DFNet-ATT Accuracy 85.2 ± 0.3 # 22


No methods listed for this paper. Add relevant methods here