Skip to content

Latest commit

 

History

History
53 lines (33 loc) · 1.39 KB

README.md

File metadata and controls

53 lines (33 loc) · 1.39 KB
Raw

Performative Time-Series Forecasting

Publication

Implementation of the paper "Performative Time-Series Forecasting."

Authors: Leo Zhiyuan Zhao, Alexander Rodríguez, B.Aditya Prakash

Paper + Appendix: https://arxiv.org/abs/2310.06077

Training

METR-LA Dataset download link: https://zenodo.org/record/5146275/files/METR-LA.csv?download=1

Example to run the covid dataset:

python3 run_covid.py --seed 0 --dev cuda:0 --model rnn --fps 0

Example to run the metrla dataset:

python3 run_metrla.py --seed 0 --dev cuda:0 --model rnn --fps 0

Example to run the metrla dataset out-of-distribution test:

python3 run_metrla_ood.py --seed 0 --dev cuda:0 --model rnn --fps 0

Implemented Models: {rnn, lstnet, transformer, informer}

fps=0 is running conventional forecasting models, fps=1 is running fps.

Contact

If you have any questions about the code, please contact Leo Zhiyuan Zhao at leozhao1997[at]gatech[dot]edu.

Citation

If you find our work useful, please cite our work:

@article{zhao2023performative,
  title={Performative Time-Series Forecasting},
  author={Zhao, Zhiyuan and Rodriguez, Alexander and Prakash, B Aditya},
  journal={arXiv preprint arXiv:2310.06077},
  year={2023}
}