Temporal convolutional autoencoder for unsupervised anomaly detection in time series

Applied Soft Computing, vol. 112, pp. 107751

Abstract

Learning temporal patterns in time series remains a challenging task up until today. Particularly for anomaly detection in time series, it is essential to learn the underlying structure of a system’s normal behavior. Periodic or quasiperiodic signals with complex temporal patterns make the problem even more challenging: Anomalies may be a hard-to-detect deviation from the normal recurring pattern. In this paper, we present TCN-AE, a t emporal c onvolutional n etwork a uto e ncoder based on dilated convolutions. Contrary to many other anomaly detection algorithms, TCN-AE is trained in an unsupervised manner. The algorithm demonstrates its efficacy on a comprehensive real-world anomaly benchmark comprising electrocardiogram (ECG) recordings of patients with cardiac arrhythmia . TCN-AE significantly outperforms several other unsupervised state-of-the-art anomaly detection algorithms. Moreover, we investigate the contribution of the individual enhancements and show that each new ingredient improves the overall performance on the investigated benchmark.

Authors 4

1. Markus Thill corresponding

   TH Köln - University of Applied Sciences

   Affiliation as printed

   TH Köln – University of Applied Sciences, 51643 Gummersbach, Germany

2. Wolfgang Konen

   TH Köln - University of Applied Sciences

   Affiliation as printed

   TH Köln – University of Applied Sciences, 51643 Gummersbach, Germany

3. Hao Wang Aachen

   Leiden University

   Affiliation as printed

   Leiden University, LIACS, 2333 CA Leiden, The Netherlands

4. Thomas Bäck Aachen

   Leiden University

   Affiliation as printed

   Leiden University, LIACS, 2333 CA Leiden, The Netherlands