Memory-Based Graph Networks

Abstract

Graph neural networks (GNNs) are a class of deep models that operate on data with arbitrary topology represented as graphs. We introduce an efficient memory layer for GNNs that can jointly learn node representations and coarsen the graph. We also introduce two new networks based on this layer: memory-based GNN (MemGNN) and graph memory network (GMN) that can learn hierarchical graph representations. The experimental results shows that the proposed models achieve state-of-the-art results in eight out of nine graph classification and regression benchmarks. We also show that the learned representations could correspond to chemical features in the molecule data. Code and reference implementations are released at: https://github.com/amirkhas/GraphMemoryNet

arXiv commentICLR 2020

References in corpus (13)

• Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
• Prototypical Networks for Few-shot Learning
• Neural Message Passing for Quantum Chemistry
• Graph Neural Networks: A Review of Methods and Applications
• End-To-End Memory Networks
• Discriminative Embeddings of Latent Variable Models for Structured Data
• Clustering with Deep Learning: Taxonomy and New Methods
• On Valid Optimal Assignment Kernels and Applications to Graph Classification
• Edge Contraction Pooling for Graph Neural Networks
• Neural Episodic Control
• Large Memory Layers with Product Keys
• Scaling Memory-Augmented Neural Networks with Sparse Reads and Writes
• Clique pooling for graph classification

Cited by in corpus (7)

• Contrastive Multi-View Representation Learning on Graphs
• Graph-Based Deep Learning for Medical Diagnosis and Analysis: Past, Present and Future
• Graph Neural Networks: Taxonomy, Advances and Trends
• Rethinking pooling in graph neural networks
• CoSimGNN: Towards Large-scale Graph Similarity Computation
• ChebLieNet: Invariant Spectral Graph NNs Turned Equivariant by Riemannian Geometry on Lie Groups
• CNN with large memory layers