Skip to content

Latest commit

 

History

History
32 lines (22 loc) · 981 Bytes

README.md

File metadata and controls

32 lines (22 loc) · 981 Bytes

Graph Neural Networks Beyond Compromise Between Attribute and Topology

This repository contains a PyTorch implementation of "Graph Neural Networks Beyond Compromise Between Attribute and Topology".

Runtime Environment

  • python 3.8
  • numpy 1.20.3
  • pytorch 1.8.0
  • pytorch-sparse 0.6.11

Dataset Source

All datasets are downloaded from package torch_geometric and saved as series of .pt file without any preprocess procedure. You can download the zipped dataset from release page of this repo and extract them to "%PROJECT_ROOT%/dataset" folder.

Run All Benchmarks

./train.sh

Citation

@article{yangWWW2022gnnbc,
  title = {Graph Neural Networks Beyond Compromise Between Attribute and Topology},
  author = {Liang Yang, Wenmiao Zhou and Weihang Peng, Bingxin Niu and Junhua Gu, Chuan Wang and Xiaochun Cao, Dongxiao He},
  year = {2022},
  booktitle = {{WWW} '22: The {ACM} Web Conference 2022},
}