Sequential Anomaly Detection with Local and Global Explanations (GLEAD)

GLEAD: Sequential Anomaly Detection with Local and Global Explanations

We propose GLEAD, a framework for sequential anomaly detection that provides both local explanations (highlighting anomalous subsequences within a sequence) and global explanations (characterizing patterns across datasets). This dual perspective improves transparency for practitioners while preserving strong detection performance. Experiments on system log datasets demonstrate that GLEAD yields interpretable insights and competitive anomaly detection accuracy.

Citation

@inproceedings{cheng2022sequential,
  title={Sequential anomaly detection with local and global explanations},
  author={Cheng, He and Xu, Depeng and Yuan, Shuhan},
  booktitle={2022 IEEE International Conference on Big Data (Big Data)},
  pages={1212--1217},
  year={2022},
  organization={IEEE}
}