Time Transformer

Time-Transformer

Pytnon implementation paper "Time-Transformer: Integrating Local and Global Features for Better Time Series Generation" (SDM24).

Jupyter Notebook "tutorial" provide a tutorial for training and evaluating with different metrics (using "sine_cpx" dataset). FID score are calculated with "fid_score" in ts2vec, directly using model "TS2Vec".

The model is built with "tensorflow2", please check the "requirement.txt" and decide which package you need to run the model.

If you find this model useful and put it in your publication, we encourage you to add the following references:

@inproceedings{liu2024time,
  title={Time-Transformer: Integrating Local and Global Features for Better Time Series Generation},
  author={Liu, Yuansan and Wijewickrema, Sudanthi and Li, Ang and Bester, Christofer and O'Leary, Stephen and Bailey, James},
  booktitle={Proceedings of the 2024 SIAM International Conference on Data Mining (SDM)},
  pages={325--333},
  year={2024},
  organization={SIAM}
}