Application Performance Monitoring Now Enables Machine Learning-Based Anomaly Detection

Anomaly detection is now enabled for custom Metric Groups. You can now use the Oracle Cloud Infrastructure Monitoring service to view and obtain alerts on anomaly detection events.

A Machine Learning model is calculated for any custom metric defined in a Metric Group. A new dimension, Anomaly, is added to each of the metrics. The dimension values are:

  • 1 – higher value anomaly detected
  • 0 – no anomaly detected
  • -1 – lower value anomaly detected

For information about Metric Groups, see Application Performance Monitoring documentation.