Distributed Explainable Ensemble Anomaly Detection for Cloud-Native Applications Using Azure AI and SQL Analytics
International Journal of Technology, Management and Humanities · 15.01.2025 · 11 (01) · 54-58
DOI: 10.21590/ijtmh.11.01.07 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal of Technology, Management and Humanities, online erschienen am 15.01.2025. Verfasst von George Zacharia.
Abstract (Originalsprache)
In this study, we propose an ensemble anomaly detection architecture based on explainability and distributed computing. Using a combination of Azure Machine Learning and distributed SQL-based analytics, our method attempts to handle large amounts of telemetry data while remaining fully interpretable throughout the process. Our proposed framework involves the combination of three models, including Isolation Forest, LightGBM, and a time-series detector using CNN, and utilizes the outputs from these three in coordination by means of Azure Synapse Analytics, which allows us to adjust the weightings dynamically. To maintain the explainability of our pipeline, we used game-theoretic Shapley values to attribute anomalies to particular features and even sub-models in near real-time.
Zitieren
APA
Zacharia, G. (2025). Distributed Explainable Ensemble Anomaly Detection for Cloud-Native Applications Using Azure AI and SQL Analytics. International Journal of Technology, Management and Humanities, 11(01), 54-58. https://doi.org/10.21590/ijtmh.11.01.07
Vancouver
Zacharia G. Distributed Explainable Ensemble Anomaly Detection for Cloud-Native Applications Using Azure AI and SQL Analytics. International Journal of Technology, Management and Humanities. 2025;11(01):54-58. doi:10.21590/ijtmh.11.01.07
BibTeX
@article{Zacharia2025,
author = {Zacharia, George},
title = {Distributed Explainable Ensemble Anomaly Detection for Cloud-Native Applications Using Azure AI and SQL Analytics},
journal = {International Journal of Technology, Management and Humanities},
year = {2025},
volume = {11},
number = {01},
pages = {54--58},
doi = {10.21590/ijtmh.11.01.07}
}
EUVIMED