Anomaly detection is a statistical approach for identifying odd data points or events that depart from the norm. It is employed to distinguish “true signals” from “noise” and pinpoint possible anomaly causes. Outlier detection is another name for anomaly detection.