Edge Contraction Pooling for Graph Neural Networks

Abstract

Graph Neural Network (GNN) research has concentrated on improving convolutional layers, with little attention paid to developing graph pooling layers. Yet pooling layers can enable GNNs to reason over abstracted groups of nodes instead of single nodes. To close this gap, we propose a graph pooling layer relying on the notion of edge contraction: EdgePool learns a localized and sparse hard pooling transform. We show that EdgePool outperforms alternative pooling methods, can be easily integrated into most GNN models, and improves performance on both node and graph classification.

References in corpus (6)

• Adam: A Method for Stochastic Optimization
• Inductive Representation Learning on Large Graphs
• Fast Graph Representation Learning with PyTorch Geometric
• Pitfalls of Graph Neural Network Evaluation
• Self-Attention Graph Pooling
• Towards Sparse Hierarchical Graph Classifiers

Cited by in corpus (23)

• Hierarchical Graph Pooling with Structure Learning
• Understanding Pooling in Graph Neural Networks
• Accurate Learning of Graph Representations with Graph Multiset Pooling
• Hierarchical Representation Learning in Graph Neural Networks with Node Decimation Pooling
• Graph Pooling via Coarsened Graph Infomax
• Higher-order Clustering and Pooling for Graph Neural Networks
• Implementing graph neural networks with TensorFlow-Keras
• Cross-Modal Hierarchical Modelling for Fine-Grained Sketch Based Image Retrieval
• Edge Representation Learning with Hypergraphs
• Memory-Based Graph Networks
• Recognizing Predictive Substructures with Subgraph Information Bottleneck
• Adaptive-Step Graph Meta-Learner for Few-Shot Graph Classification
• MxPool: Multiplex Pooling for Hierarchical Graph Representation Learning
• LookHops: light multi-order convolution and pooling for graph classification
• A Dynamic Reduction Network for Point Clouds
• Using graph neural networks to reconstruct charged pion showers in the CMS High Granularity Calorimeter
• Graph Information Bottleneck for Subgraph Recognition
• Learning Hierarchical Review Graph Representations for Recommendation
• Embedding Graphs on Grassmann Manifold
• Quadratic GCN for Graph Classification
• GeoGraph: Learning graph-based multi-view object detection with geometric cues end-to-end
• Edge but not Least: Cross-View Graph Pooling
• Automatic design of novel potential 3CL${}^{\text{pro}}$ and PL${}^{\text{pro}}$ inhibitors