高效可扩展图神经网络的研究进展，Recent Advances in Efficient and Scalable Graph Neural Networks

这篇文章的目的是提供一个关于高效的图神经网络和可扩展的图表示学习的关键思想的概述。我们将涵盖数据准备、GNN架构和学习范例的关键发展，使图神经网络能够扩展到真实世界的图形和实时应用程序。

🚀 Efficient Graph Neural Networks

• Real-world Challenges for Graph Neural Networks
  1. Giant Graphs – Memory Limitations
  2. Sparse Computations – Hardware Limitations
  3. Graph Subsampling – Reliability Limitations
• Handling Large-scale Graphs
  • Graph Subsampling Techniques
  • Historical Node Embeddings
• Scalable and Resource-efficient GNN Architectures
  • Graph-augmented MLPs
  • Efficient Graph Convolutional Layers Learning Paradigms for GNN Compression
  • Knowledge Distillation for Boosting Performance
  • Quantization for Low Precision GNNs Conclusion and Outlook Comprehensive Surveys