Real time anomaly detection and categorisation

Abstract The ability to quickly and accurately detect anomalous structure within data sequences is an inference challenge of growing importance. This work extends recently proposed post-hoc (offline) anomaly detection methodology the sequential setting. resultant procedure capable real-time analysis categorisation between baseline two forms structure: point collective anomalies. Various theoretical properties are derived. These, together with extensive simulation study, highlight that average run length false alarm delay online algorithm very close offline version. Experiments on simulated real provided demonstrate benefits method.