XEM: An explainable-by-design ensemble method for multivariate time series classification

• Published in: Data mining and knowledge discovery Vol. 36; no. 3; pp. 917 - 957
• Main Authors: Fauvel, Kevin, Fromont, Élisa, Masson, Véronique, Faverdin, Philippe, Termier, Alexandre
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.05.2022; Springer Nature B.V; Springer
• Subjects: Artificial Intelligence; Chemistry and Earth Sciences; Classification; Classifiers; Computer Science; Data collection; Data Mining and Knowledge Discovery; Information Storage and Retrieval; Machine learning; Missing data; Multivariate analysis; Physics; Statistics for Engineering; Time series; Multivariate time series; Ensemble learning; Explainability; Classification; Multivariate Time Series; Ensemble Learning
• ISSN: 1384-5810; 1573-756X
• DOI: 10.1007/s10618-022-00823-6

We present XEM, an eXplainable-by-design Ensemble method for Multivariate time series classification. XEM relies on a new hybrid ensemble method that combines an explicit boosting-bagging approach to handle the bias-variance trade-off faced by machine learning models and an implicit divide-and-conquer approach to individualize classifier errors on different parts of the training data. Our evaluation shows that XEM outperforms the state-of-the-art MTS classifiers on the public UEA datasets. Furthermore, XEM provides faithful explainability-by-design and manifests robust performance when faced with challenges arising from continuous data collection (different MTS length, missing data and noise).