Reference: https://docs.splunk.com/Documentation/ITSI/4.10.2/SI/AD
Anomaly detection is a feature of ITSI that uses machine learning to detect when KPI data deviates
from a normal pattern. The following items apply to anomaly detection:
B) A minimum of 24 hours of data is needed for anomaly detection, and a minimum of 4 entities for
cohesive analysis. This ensures that there is enough data to establish a baseline pattern and compare
different entities within a service.
C) Anomaly detection automatically generates notable events when KPI data diverges from the
pattern. You can configure the sensitivity and severity of the anomaly detection alerts and assign
them to episodes or teams. Reference: [Anomaly Detection]
