/ST-SSL

ST-SSL (STSSL): Spatio-Temporal Self-Supervised Learning for Traffic Flow Forecasting/Prediction

Primary LanguagePython

ST-SSL: Spatio-Temporal Self-Supervised Learning for Traffic Prediction

PWCPWCPWCPWC

This is a Pytorch implementation of ST-SSL in the following paper:

  • J. Ji, J. Wang, C. Huang, et al. "Spatio-Temporal Self-Supervised Learning for Traffic Flow Prediction". in Thirty-Seventh AAAI Conference on Artificial Intelligence, 2023.

The homepage of J. Ji is available at here.

Requirement

We build this project by Python 3.8 with the following packages:

numpy==1.21.2
pandas==1.3.5
PyYAML==6.0
torch==1.10.1

Model training and Evaluation

If the environment is ready, please run the following commands to train model on the specific dataset from {NYCBike1, NYCBike2, NYCTaxi, BJTaxi}.

>> cd ST-SSL
>> ./runme 0 NYCBike1   # 0 gives the gpu id

This repo contains the NYCBike1 data. If you are interested in other datasets, please download from ST-SSL_Dataset.

Cite

If you find the paper useful, please cite as following:

@inproceedings{ji2023spatio, 
  title={Spatio-Temporal Self-Supervised Learning for Traffic Flow Prediction}, 
  author={Ji, Jiahao and Wang, Jingyuan and Huang, Chao and Wu, Junjie and Xu, Boren and Wu, Zhenhe and Zhang, Junbo and Zheng, Yu}, 
  booktitle={Proceedings of the AAAI Conference on Artificial Intelligence}, 
  year={2023}
}