In an analytics search, you can query the CMDB by clicking Value > Select from CMDB, which allows you to choose values directly from CMDB entries for the selected attribute, enabling precise filtering based on asset data.

The ML regression model uses memory utilization, sent bytes, and received bytes as prediction inputs for CPU utilization. If memory usage remains consistently high, FortiSIEM adapts the learned baseline and updates the model with a higher average memory utilization value over time.

FortiSIEM can retrieve ZTNA tags from FortiClient EMS through an API connection, enabling dynamic user and device classification for policy enforcement and incident response.

A Regression algorithm is used when predicting a continuous or numeric target field based on other features in the dataset. In FortiSIEM, regression-based machine learning models establish expected values, and an anomaly is triggered when the actual observed value significantly deviates from the regression prediction.

FortiSIEM dashboard widgets display data based on reports. Each widget uses an underlying report or analytics query to define the dataset, aggregation, and visualization shown on the dashboard.