Towards Generalizable Graph Contrastive Learning: An Information Theory Perspective

Yige Yuan, Bingbing Xu, Huawei Shen, Qi Cao, Keting Cen, Wen Zheng, Xueqi Cheng

Neural Networks (NN), Volume 172, April 2024

This is an official PyTorch implementation of paper Towards Generalizable Graph Contrastive Learning: An Information Theory Perspective.

Requirements

To install requirements:

pip install -r requirements.txt

Training & Evaluating

To train the model(s) in the paper, run this command:

python -u train.py --dataset Cora  --if_save True  --save_path ./log

Reference

If you find our work useful, please consider citing our paper:

@article{yuan2022towards,
  title={Towards Generalizable Graph Contrastive Learning: An Information Theory Perspective},
  author={Yuan, Yige and Xu, Bingbing and Shen, Huawei and Cao, Qi and Cen, Keting and Zheng, Wen and Cheng, Xueqi},
  journal={arXiv preprint arXiv:2211.10929},
  year={2022}
}