Shapelet-based Model-agnostic Counterfactual Local Explanations for Time Series Classification
Machine Learning
2024-02-05 v1
Abstract
In this work, we propose a model-agnostic instance-based post-hoc explainability method for time series classification. The proposed algorithm, namely Time-CF, leverages shapelets and TimeGAN to provide counterfactual explanations for arbitrary time series classifiers. We validate the proposed method on several real-world univariate time series classification tasks from the UCR Time Series Archive. The results indicate that the counterfactual instances generated by Time-CF when compared to state-of-the-art methods, demonstrate better performance in terms of four explainability metrics: closeness, sensibility, plausibility, and sparsity.
Cite
@article{arxiv.2402.01343,
title = {Shapelet-based Model-agnostic Counterfactual Local Explanations for Time Series Classification},
author = {Qi Huang and Wei Chen and Thomas Bäck and Niki van Stein},
journal= {arXiv preprint arXiv:2402.01343},
year = {2024}
}
Comments
The paper has been accepted by the XAI4Sci workshop of AAAI 2024