A Decade Review: Diving Deep into Time-Series Anomaly Detection

Advances in data collection and the explosion of streaming data highlight the crucial need for time-series analytics. This paper provides a decade-long review of time-series anomaly detection, encompassing methods from traditional statistical measures to the surge of machine learning algorithms. It presents a process-centric taxonomy to categorize and summarize existing solutions, offering a meta-analysis of the literature and outlining general trends in the field. This comprehensive survey serves as a valuable resource for researchers.