tshap

This repo accompanies our ECML-PKDD 2025 research track paper TSHAP: Fast and Exact SHAP for Explaining Time Series Classification and Regression.

TSHAP is a SHAP-based attribution method for time series classification and regression. The attributions can be used to explain the prediction of the time series model. TSHAP has two variants: TSHAP-Window and TSHAP-ROI. It supports both univariate and multivariate time series data for the tasks of classification and extrinsic regression.

To use TSHAP, include the source Python module or install with Pypi.

pip install tshap

Examples can be found in the Jupyter notebook.

Citation

If you use this repository in your research, please cite as:

@misc{lenguyen2025tshap,
    title={[TSHAP: Fast and Exact SHAP for Explaining Time Series Classification and Regression},
    author={Thach Le Nguyen and Georgiana Ifrim},
    year={2025},
    conference={ECMLPKDD},
    eprint={https://www.researchgate.net/publication/392833501_TSHAP_Fast_and_Exact_SHAP_for_Explaining_Time_Series_Classification_and_Regression},
    archivePrefix={arXiv},
    primaryClass={cs.LG}
}