图形神经网络

1.什么是图形神经网络

百度百科
图网络（Graph Network, GN）是在拓扑空间（topological space）内按图（graph）结构组织以进行关系推理（relational reasoning）的函数集合。在深度学习理论中是图神经网络（ graph neural network, GNN）和概率图模型（Probabilistic Graphical Model, PGM）的推广 。
图网络由图网络块（GN block）构成，具有灵活的拓扑结构，可以特化为各类连接主义（connectionist）模型，包括前馈神经网络（Feedforward Neural Network）、递归神经网络（Recursive Neural Network）等 。更一般的图网络适用于处理具有图结构的数据，例如知识图谱、社交网络、分子网络等。

2.理解

没找到《Relational inductive biases, deep learning, and graph networks》这篇论文，下载速度太慢了，但是找到一篇博客，写的是对这篇论文的理解，还挺详细，图网络模型原理详解。图网络的节点和边的关系构造和c语言的链表有些像。

3.相关论文

1. Peter W. Battaglia Jessica B.Relational inductive biases, deep learning, and graph networks. arXiv:1806.01261（对图网络的综述）
2. A Sanchez-Gonzalez,N Heess.Graph networks as learnable physics engines for inference and control.arXiv:1806.01
3. David I Shuman, Sunil K. Narang .The emerging field of signal processing on graphs: Extending high-dimensional data analysis to networks and other irregular domains.IEEE Signal Processing Magazine ( Volume: 30 , Issue: 3 , May 2013 )
4. Michaël Defferrard,XavierBresson.Pierre Vandergheynst,Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering.Advances in Neural Information Processing Systems 29 (NIPS 2016)
5. Thomas N. Kipf, Max Welling .Semi-Supervised Classification with Graph Convolutional Networks.arXiv:1609.02907
6. P Velickovic, G Cucurull, A Casanova.Graph attention networks
7. Alexandre Alahi, Kratarth Goel.Social LSTM: Human Trajectory Prediction in Crowded Spaces.The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016, pp. 961-971
8. KiJung Yoon, Renjie Liao.Inference in Probabilistic Graphical Models by Graph Neural Networks.arXiv:1803.07710
9. Irwan Bello, Hieu Pham.Neural Combinatorial Optimization with Reinforcement Learning. arXiv:1611.09940
10. Eric Jang, Shixiang Gu, Ben Poole .Categorical Reparameterization with Gumbel-Softmax.arXiv:1611.01144
11. Aleksandar Bojchevski, Oleksandr Shchur.NetGAN: Generating Graphs via Random Walks.arXiv:1803.00816
12. Daniel D. Johnson.Learning Graphical State Transitions
13. Ashish Vaswani,Noam Shazeer.Attention is All you Need. Advances in Neural Information Processing Systems 30 (NIPS 2017)

好

4.在视频中看到的（待完善）

1.图网络的可微分性
2. 图网络进行卷积操作
3. network embedding
4.由局部到整体