COGS

COGS: A Causal Representation Learning Framework for Out-of-Distribution Generalization in Time Series

Building Conda Environment

cd COGS
conda env create -f environment.yml

conda activate cogs

Running

Prepare your own data and move it to ./data dir. Here we provide an example of DSADS dataset and you can download it from DSADS Dataset.

After data preparation, you can run the following script.

python main.py

Citations

@inproceedings{song2026cogs,
  title={COGS: A Causal Representation Learning Framework for Out-of-Distribution Generalization in Time Series},
  author={Song, Xinxin and Cheng, Yuxiao and Xiao, Tingxiong and Suo, Jinli},
  booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
  volume={40},
  number={30},
  pages={25572--25580},
  year={2026}
}