denx20 / Att-McLSTM

Official PyTorch Implementation of "Attention-based recurrent neural network for influenza epidemic prediction"

Att-McLSTM

This is an official PyTorch implementation of the Att-McLSTM model presented in the following paper:

Attention-based recurrent neural network for influenza epidemic prediction

General

Project directories:

• data: data analysis and pre-processing
• model: source code
• experiment: comparative experiments and ablation study
• utils: some useful scripts for logging and showing demos

Dependencies

This project uses Python 3.6 and Keras. I recommend using anaconda for dependency management:

conda create -n att-mclstm python=3.6
conda activate att-mclstm

Note that this requires CUDA 9.2. Depending on your cuda version, you may want to install previous versions of TensorFlow. See here.

Reproducibility

The results of all functions and training loop match the results demonstrated in the paper. One can check it out and run to test:

python ./data/sum_data.py
python ./model/att_multi_channel_lstm.py

Misc

Citation

If you find our work useful in your research, please cite:

@article{zhu2019attention,
  title={Attention-based recurrent neural network for influenza epidemic prediction},
  author={Zhu, Xianglei and Fu, Bofeng and Yang, Yaodong and Ma, Yu and Hao, Jianye and Chen, Siqi and Liu, Shuang and Li, Tiegang and Liu, Sen and Guo, Weiming and others},
  journal={BMC bioinformatics},
  volume={20},
  number={18},
  pages={1--10},
  year={2019},
  publisher={BioMed Central}
}

Contact

If you have any questions, please email Bofeng Fu at [e-mail].