EUVIMEDEUVIMED entdecken
Crossref

VUS: effective and efficient accuracy measures for time-series anomaly detection

Paul Boniol, Ashwin K. Krishna, Marine Bruel, Qinghua Liu, Mingyi Huang, Themis Palpanas, Ruey S. Tsay, Aaron Elmore, Michael J. Franklin, John Paparrizos

The VLDB Journal · 2025-05 · 34 (3)

DOI: 10.1007/s00778-025-00907-x ↗

Kurz gefasst

Fachartikel in der Zeitschrift The VLDB Journal, online erschienen am 27.03.2025. Verfasst von Paul Boniol und 9 weiteren Autor:innen. Die Arbeit wurde bisher 40-mal zitiert (Crossref) und nennt 58 Literaturangaben. Gefördert von: Grand Équipement National De Calcul Intensif, Horizon 2020, TwinsODIS.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Anomaly detection (AD) is a fundamental task for time-series analytics with important implications for the downstream performance of many applications. In contrast to other domains where AD mainly focuses on point-based anomalies (i.e., outliers in standalone observations), AD for time series is also concerned with range-based anomalies (i.e., outliers spanning multiple observations). Nevertheless, it is common to use traditional point-based information retrieval measures, such as Precision, Recall, and F-score, to assess the quality of methods by thresholding the anomaly score to mark each point as an anomaly or not. However, mapping discrete labels into continuous data introduces unavoidable shortcomings, complicating the evaluation of range-based anomalies. Notably, the choice of evaluation measure may significantly bias the experimental outcome. Despite over six decades of attention, there has never been a large-scale systematic quantitative and qualitative analysis of time-series AD evaluation measures. This paper extensively evaluates quality measures for time-series AD to assess their robustness under noise, misalignments, and different anomaly cardinality ratios. Our results indicate that measures producing quality values independently of a threshold (i.e., AUC-ROC and AUC-PR) are more suitable for time-series AD. Motivated by this observation, we first extend the AUC-based measures to account for range-based anomalies. Then, we introduce a new family of parameter-free and threshold-independent measures, volume under the surface (VUS), to evaluate methods while varying parameters. We also introduce two optimized implementations for VUS that reduce significantly the execution time of the initial implementation. Our findings demonstrate that our four measures are significantly more robust in assessing the quality of time-series AD methods.

Förderung

  • Grand Équipement National De Calcul Intensif
  • Horizon 2020
  • TwinsODIS
  • ARMADA
  • DataGENS
  • RECITALS

Zitieren

APA

Boniol, P., Krishna, A. K., Bruel, M., Liu, Q., Huang, M., Palpanas, T., Tsay, R. S., Elmore, A., Franklin, M. J., & Paparrizos, J. (2025). VUS: effective and efficient accuracy measures for time-series anomaly detection. The VLDB Journal, 34(3). https://doi.org/10.1007/s00778-025-00907-x

Vancouver

Boniol P, Krishna AK, Bruel M, Liu Q, Huang M, Palpanas T, et al. VUS: effective and efficient accuracy measures for time-series anomaly detection. The VLDB Journal. 2025;34(3). doi:10.1007/s00778-025-00907-x

BibTeX
@article{Boniol2025,
  author = {Boniol, Paul and Krishna, Ashwin K. and Bruel, Marine and Liu, Qinghua and Huang, Mingyi and Palpanas, Themis and Tsay, Ruey S. and Elmore, Aaron and Franklin, Michael J. and Paparrizos, John},
  title = {VUS: effective and efficient accuracy measures for time-series anomaly detection},
  journal = {The VLDB Journal},
  year = {2025},
  volume = {34},
  number = {3},
  doi = {10.1007/s00778-025-00907-x}
}
Kostenlos starten →