Diffusion-Convolutional Neural Networks

Abstract

We present diffusion-convolutional neural networks (DCNNs), a new model for graph-structured data. Through the introduction of a diffusion-convolution operation, we show how diffusion-based representations can be learned from graph-structured data and used as an effective basis for node classification. DCNNs have several attractive qualities, including a latent representation for graphical data that is invariant under isomorphism, as well as polynomial-time prediction and learning that can be represented as tensor operations and efficiently implemented on the GPU. Through several experiments with real structured datasets, we demonstrate that DCNNs are able to outperform probabilistic relational models and kernel-on-graph methods at relational node classification tasks.

Full paper

Cited by in corpus (58)

• Geometric deep learning: going beyond Euclidean data
• Graph Neural Network for Traffic Forecasting: A Survey
• MeshCNN: A Network with an Edge
• Convolutional Neural Network Architectures for Signals Supported on Graphs
• Improving Graph Neural Network Expressivity via Subgraph Isomorphism Counting
• Spatio-temporal graph neural networks for multi-site PV power forecasting
• Relation-Aware Entity Alignment for Heterogeneous Knowledge Graphs
• ST-GRAT: A Novel Spatio-temporal Graph Attention Network for Accurately Forecasting Dynamically Changing Road Speed
• Robust Spatial Filtering with Graph Convolutional Neural Networks
• MR-GNN: Multi-Resolution and Dual Graph Neural Network for Predicting Structured Entity Interactions
• Conditional Generation Net for Medication Recommendation
• Understanding Pooling in Graph Neural Networks
• Second-Order Pooling for Graph Neural Networks
• Relation Modeling with Graph Convolutional Networks for Facial Action Unit Detection
• Representation Learning for Natural Language Processing
• Kriging Convolutional Networks
• GDPNet: Refining Latent Multi-View Graph for Relation Extraction
• Should Graph Convolution Trust Neighbors? A Simple Causal Inference Method
• AdaGNN: Graph Neural Networks with Adaptive Frequency Response Filter
• Context-Dependent Diffusion Network for Visual Relationship Detection
• Dynamic Edge-Conditioned Filters in Convolutional Neural Networks on Graphs
• Deep Learning of High-Order Interactions for Protein Interface Prediction
• Graph Pooling via Coarsened Graph Infomax
• Beyond Low-pass Filtering: Graph Convolutional Networks with Automatic Filtering
• Graph InfoClust: Leveraging cluster-level node information for unsupervised graph representation learning
• Graph Convolutional Neural Networks with Node Transition Probability-based Message Passing and DropNode Regularization
• Mining Point Cloud Local Structures by Kernel Correlation and Graph Pooling
• RawlsGCN: Towards Rawlsian Difference Principle on Graph Convolutional Network
• Higher-order Clustering and Pooling for Graph Neural Networks
• Adaptive Diffusions for Scalable Learning over Graphs
• Bridging the Gap between Spatial and Spectral Domains: A Survey on Graph Neural Networks
• Neural graph embeddings as explicit low-rank matrix factorization for link prediction
• Exploring Edge Disentanglement for Node Classification
• Bayesian graph convolutional neural networks via tempered MCMC
• Cell Attention Networks
• Hierarchical stochastic graphlet embedding for graph-based pattern recognition
• Bayesian Semi-supervised Learning with Graph Gaussian Processes
• Addressing Heterophily in Node Classification with Graph Echo State Networks
• Incomplete Graph Representation and Learning via Partial Graph Neural Networks
• Walk-Steered Convolution for Graph Classification
• SWAG: Item Recommendations using Convolutions on Weighted Graphs
• Bilinear Graph Neural Network with Neighbor Interactions
• Distance Metric Learning for Graph Structured Data
• Motion Guided 3D Pose Estimation from Videos
• Residual Correction in Real-Time Traffic Forecasting
• Learning Deep Graph Representations via Convolutional Neural Networks
• Learning to Multitask
• SpecNet: Spectral Domain Convolutional Neural Network
• Graph Learning-Convolutional Networks
• Graph Classification via Deep Learning with Virtual Nodes
• Weisfeiler and Lehman Go Paths: Learning Topological Features via Path Complexes
• How effective are Graph Neural Networks in Fraud Detection for Network Data?
• Quantum Walk Inspired Neural Networks for Graph-Structured Data
• Modeling Edge Features with Deep Bayesian Graph Networks
• Revisiting convolutional neural network on graphs with polynomial approximations of Laplace-Beltrami spectral filtering
• Where Did the President Visit Last Week? Detecting Celebrity Trips from News Articles
• Multiple Graph Adversarial Learning
• Learning Chebyshev Basis in Graph Convolutional Networks for Skeleton-based Action Recognition