Contextual Anomaly Detection

Contextual Anomaly Detection aims to identify rare and unseen objects or events in data that includes both behavioral attributes and contextual attributes. Behavioral attributes are directly related to the process being monitored, while contextual attributes involve external factors that significantly impact the process, with behavioral attributes typically depending on contextual attributes. This method, operating within an unsupervised learning framework, effectively detects anomalies by capturing the complex relationships between behavior and context, making it highly valuable for various applications.