Year: 2025
Type: conference-paper
Abstract: Time series anomaly detection plays a critical role across domains from industrial monitoring to cybersecurity use cases. But its evaluation remains challenging. Traditional window level F1 score overweights long anomaly intervals, while heuristic “point‑adjusted” variants introduce bias by extending single detection across entire zones. We propose a Zone Normalized F1, which treats each true and each predicted anomaly interval as a unit, macro‑averaging precision and recall over intervals rather than windows. This eliminates length bias and yields a fairer comparison of detectors. We formalize the metric, illustrate its behavior on toy and real examples, and show how it complements existing protocols.
Authors: Nabil Ait Said, Elies Gherbi, Faouzi Adjed, Achraf Kallel
Institution: Institut de Recherche Technologique SystemX
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W4416166275
FWCI:
Cites:
Cited by:
Related to: