@misc{10498/39273,
year = {2025},
url = {http://hdl.handle.net/10498/39273},
abstract = {The time series classification problem is highly relevant in many important domains, such as industry, finance, and healthcare. It is a long-known problem that has been addressed by a plethora of mathematical algorithms, including deep learning models. Despite the rapid growth of deep learning techniques, the opaque nature of these models prevents users from adopting them in real-life scenarios. In this paper, we propose a classification framework for time series using wavelet transformation and transformers. Furthermore, we explain the results using a novel combination of the attention mechanism and Linear Discriminant Analysis. The framework was evaluated on three real-world wind turbine datasets, and it achieved comparable or leading F1 scores of above 0.99. Moreover, the explanation framework highlights the key frequencies that differentiate between the healthy and faulty states. We show that these frequencies align by at least 70% with those identified using mechanical domain knowledge. This compliance increases the trustworthiness of Artificial Intelligence in industrial applications and supports informed decision-making.},
keywords = {Artificial intelligence},
keywords = {attention mechanisms},
keywords = {classification},
keywords = {explainable artificial intelligence},
keywords = {fault detection},
keywords = {linear discriminant analysis},
keywords = {machine learning},
keywords = {transformer},
keywords = {wavelet transforms},
keywords = {wind turbines},
title = {Explainable Artificial Intelligence for Time Series Using Attention Mechanism: Application to Wind Turbine Fault Detection},
doi = {10.1109/ACCESS.2025.3621003},
author = {Ghahfarokhi, Anahita Farhang and Schäfer, Jörg and Wagner, Matthias F. and Dorronsoro Díaz, Bernabé},
}