U-GCN

This repository contains the demo code of the paper:

Universal Graph Convolutional Networks

which has been accepted by NeurIPS2021.

Dependencies

• Python3
• NumPy
• SciPy
• NetworkX
• PyTorch

Example

bash run_data.sh

Acknowledgements

The demo code is implemented based on AM-GCN: Adaptive Multi-channel Graph Convolutional Networks.

Citing

@inproceedings{U-GCN,
 title={Universal Graph Convolutional Networks},
 author={Di Jin, Zhizhi Yu, Cuiying Huo, Rui Wang, Xiao Wang, Dongxiao He, and Jiawei Han},
 booktitle = {NeurIPS},
 year={2021}
}