• EN

Efficient Subsequence Anomaly Detection On Time Series Data

We are investigating algorithms, that can detect anomalous subsequences on time series. An anomaly can stem from different events, such as errors in sensors or constructions, diseases, or special events. Many challenges arise in the process of detecting anomalies:

  • Length of a time series
  • Width of a time series
  • Anomaly contribution of different dimensions
  • Generalization to different data characteristics

Subprojects

  • AutoTSAD: Unsupervised System to Automatically Select and Configure Time Series Anomaly Detection Algorithms
  • CorrA: Correlation Anomaly Detection in High-Dimensional Time Series
  • JET: Fast Estimation of Hierarchical Time Series Clustering
  • HYPEX: Hyperparameter Optimization in Time Series Anomaly Detection
  • CAST: Classifying Anomalous Subsequences in Time Series
  • Series2Graph++: Distributed Detection of Correlation Anomalies in Multivariate Time Series
  • Anomaly Detection in Time Series: A Comprehensive Evaluation
  • DADS: Distributed Detection of Sequential Anomalies in Univariate Time Series

Publications

  • Anthony Bagnall, Matthew Middlehurst, Germain Forestier, Ali Ismail-Fawaz, Antoine Guillaume, David Guijo-Rubio, Arik Ermshaus, Patrick Schäfer, Thorsten Papenbrock, Phillip Wenig, Sebastian Schmidl: An Introduction to Machine Learning from Time Series. Proceedings of the European Conference on Machine Learning and Data Mining (ECML PKDD), 2024 (to appear)
  • Sebastian Schmidl, Naumann Felix, Papenbrock Thorsten: AutoTSAD: Unsupervised Holistic Anomaly Detection for Time Series Data. PVLDB 17:(11), 2024
    [Paper]  [vldb]  [Project Page]  [DOI:10.14778/3681954.3681978]
  • Phillip Wenig, Sebastian Schmidl, Thorsten Papenbrock: Anomaly Detectors for Multivariate Time Series: The Proof of the Pudding is in the Eating. Proceedings of the International Conference on Data Engineering Workshops (ICDEW), 2024
    [Paper]  [DOI:10.1109/ICDEW61823.2024.00018]
  • Sebastian Schmidl, Phillip Wenig, Thorsten Papenbrock: HYPEX: Hyperparameter Optimization in Time Series Anomaly Detection. Proceedings of the conference on Database Systems for Business, Technology, and Web (BTW), 2023
    [Paper]  [Project Page]  [DOI:10.18420/BTW2023-22]
  • Phillip Wenig, Sebastian Schmidl, Thorsten Papenbrock: TimeEval: A Benchmarking Toolkit for Time Series Anomaly Detection Algorithms. PVLDB 12:(15), 2022
    [Paper]  [Project Page]  [DOI:10.14778/3554821.3554873]
  • Sebastian Schmidl, Phillip Wenig, Thorsten Papenbrock: Anomaly Detection in Time Series: A Comprehensive Evaluation. PVLDB 9:(15), 2022
    [Paper]  [Poster]  [Project Page]  [DOI:10.14778/3538598.3538602]
  • Johannes Schneider, Phillip Wenig, Thorsten Papenbrock: Distributed detection of sequential anomalies in univariate time series. The VLDB Journal (2021)
    [Paper]  [Poster]  [Project Page]  [DOI:10.1007/s00778-021-00657-6]