Anomaly detection Machine Learning
noun phrase
Definition: The task of identifying data points, events, or patterns that deviate substantially from expected behavior or the majority of observations [Google ML Glossary; scikit-learn Documentation].
Example in context: “Addressing these research questions can advance anomaly detection in critical infrastructure systems, enhancing their resilience and security against adversarial threats.” [Cali et al. 2024]
Synonym: outlier detection
Related terms: novelty detection, fraud detection