Anomaly Detection

Anomaly detection is an ML task focused on identifying patterns or instances that deviate significantly from the normal behavior of a system. The goal is to automatically detect rare or abnormal observations that might indicate potential anomalies, outliers, or unusual events.

Anomaly detection plays a crucial role in various domains such as fraud detection, network security, system monitoring, manufacturing quality control, and healthcare. It helps in detecting unusual or potentially harmful events that might require further investigation or mitigation, providing valuable insights and early warnings for maintaining system integrity and reliability.

Anomaly detection application structure

ML.NET doesn’t have a dedicated CLI command for anomaly detection; therefore, we need to write our own code. We've done so in the playground below. For the convenience of demonstration, we placed most of the logic into the Program.cs file.