Machine Learning Comparative Reporting
Machine Learning (ML) Comparative Reporting enables you to compare ML anomaly results with Validation, Editing, and Estimation (VEE) results and other corrective actions. You can use this functionality to evaluate anomaly score thresholds before deployment and to monitor the effectiveness of ML anomaly scoring in production.
Before configuring comparative reporting, configure Machine Learning Anomaly Scoring, including the required machine learning model, payload processing, and batch controls. This includes set up and configuration of Oracle Cloud Object Storage Buckets (see Creating Object Storage Locations for more information).
Machine Learning (ML) Comparative Reporting can be used in two ways:
-
Historical Evaluation evaluates existing interval measurement data (IMDs) processed using VEE for a selected period.
-
Ongoing Evaluation captures comparison statistics for ML exceptions created during normal processing.