Back to Floor
Anomaly Detection & Ops • Room 1
Anomaly Detection
ITSI's Anomaly Detection uses machine learning to identify unusual KPI patterns that static thresholds might miss. It analyzes historical trends and flags deviations from expected behavior.
Anomaly Detection is especially useful for KPIs with complex seasonal patterns. Retail transaction volume differs on weekdays vs. weekends, holidays vs. normal days. ML learns these patterns and only alerts on true deviations.
Anomaly Detection needs 2-4 weeks of historical data minimum. Don't enable it on a brand-new KPI.
Knowledge Check
Prove your understanding to clear the room (Rewards XP)
Question 1 of 2
How much historical data does Anomaly Detection typically need?