- Title
- A review of time-series anomaly detection techniques: a step to future perspectives
- Creator
- Shaukat, Kamran; Alam, Talha Mahboob; Luo, Suhuai; Shabbir, Shakir; Hameed, Ibrahim A.; Li, Jiaming; Abbas, Syed Konain; Javed, Umair
- Relation
- Advances in Information and Communication: Proceedings of the 2021 Future of Information and Communication Conference (FICC) p. 865-877
- Relation
- Advances in Intelligent Systems and Computing Vol. 1363
- Publisher Link
- http://dx.doi.org/10.1007/978-3-030-73100-7_60
- Publisher
- Springer Cham
- Resource Type
- book chapter
- Date
- 2021
- Description
- Anomaly detection is a significant problem that has been studied in a broader spectrum of research areas due to its diverse applications in different domains. Despite the usage of modern technologies and the advances in system monitoring and anomaly detection techniques, false-positive rates are still high. There exist many anomaly detection algorithms, among them few are domain-specific, and others are more generic techniques. Despite a significant amount of advance in this research area, there does not exist a single winning anomaly detector known to work well across different datasets. In this paper, we review the literature related to types of anomalies, data types of anomalies, data types of time-series, components of time-series data, classification of anomalies context, and classification methods of time-series anomalies detection. We presented a taxonomy to characterize the various aspects related to time-series anomaly detection. One of the key challenges in current anomaly detection techniques is to perform anomaly detection with regards to the type of activities or the context that a system is exposed. We hope that this investigation gives a more remarkable ability to understand the evolving methods of time-series anomaly detection and how computational methods can be applied in this domain in the future.
- Subject
- anomaly detection; anomalies context; anomaly types; anomaly data typs; machine learning; review; time-series
- Identifier
- http://hdl.handle.net/1959.13/1462111
- Identifier
- uon:46384
- Identifier
- ISBN:9783030730994
- Language
- eng
- Hits: 2030
- Visitors: 2020
- Downloads: 0
| Thumbnail | File | Description | Size | Format |
|---|