Adaptive Graph Convolutional Neural Networks

Abstract

Graph Convolutional Neural Networks (Graph CNNs) are generalizations of classical CNNs to handle graph data such as molecular data, point could and social networks. Current filters in graph CNNs are built for fixed and shared graph structure. However, for most real data, the graph structures varies in both size and connectivity. The paper proposes a generalized and flexible graph CNN taking data of arbitrary graph structure as input. In that way a task-driven adaptive graph is learned for each graph data while training. To efficiently learn the graph, a distance metric learning is proposed. Extensive experiments on nine graph-structured datasets have demonstrated the superior performance improvement on both convergence speed and predictive accuracy.

The Thirty-Second AAAI Conference on Artificial Intelligence (AAAI-18), 8 pages

References in corpus (5)

• Semi-Supervised Classification with Graph Convolutional Networks
• Deep Convolutional Networks on Graph-Structured Data
• Atomic Convolutional Networks for Predicting Protein-Ligand Binding Affinity
• Deep Learning with Dynamic Computation Graphs
• Low Data Drug Discovery with One-shot Learning

Cited by in corpus (9)

• Enhancing Graph Neural Network-based Fraud Detectors against Camouflaged Fraudsters
• DeeperGCN: All You Need to Train Deeper GCNs
• Hierarchical Graph Pooling with Structure Learning
• TransGCN:Coupling Transformation Assumptions with Graph Convolutional Networks for Link Prediction
• DefenseVGAE: Defending against Adversarial Attacks on Graph Data via a Variational Graph Autoencoder
• Stability and Generalization of Graph Convolutional Neural Networks
• LaplacianNet: Learning on 3D Meshes with Laplacian Encoding and Pooling
• Graph Attribute Aggregation Network with Progressive Margin Folding
• Multiple Graph Adversarial Learning