This is an implementation of Traffic Inflow and Outflow Forecasting by Modeling Intra- and Inter-Relationship Between Flows (TITS). MR-STN is a novel deep spatio-temporal network framework for traffic inflows and outflows forecasting. We show the generality and superiority of MR-STN by implementing it with four state-of-the-art graph-based deep spatio-temporal models, including STGCN, ASTGCN, STMGCN, and STSGCN.

• mxnet>=1.5.0
• easydict

Use nvcc -V to check the cuda version and install mxnet with the corresponding version. For example, use pip install mxnet-cu101 to install mxnet for cuda version 10.1.

Please download the data and unzip it in the ./dataset directory.

• python main.py --rid=1 --mode=stgcn --stack=3 --ed=4 --data=Metro
• python main.py --rid=1 --mode=astgcn --stack=3 --ed=4 --data=Metro
• python main.py --rid=1 --mode=stmgcn --stack=3 --ed=4 --data=Metro
• python main.py --rid=1 --mode=stsgcn --stack=3 --ed=4 --data=Metro

If our paper benefits to your research, please cite our paper using the bitex below:

@article{MRSTN,
    title={Traffic Inflow and Outflow Forecasting by Modeling Intra- and Inter-Relationship Between Flows},
    author={Zhao, Yiji and Lin, Youfang and Zhang, Yongkai and Wen, Haomin and Liu, Yunxiao and Wu, Hao and Wu, Zhihao and Zhang, Shuaichao and Wan, Huaiyu},
    journal={IEEE Transactions on Intelligent Transportation Systems},
    volume={23},
    number={11},
    pages={20202--20216},
    year={2022},
    publisher={IEEE}
}