Detect outliers and anomalies

Anomaly detection finds unexpected time-series values

Anomaly detection works on supported line-chart time series. Power BI calculates an expected range and marks points that deviate from that range.

Sensitivity changes what is flagged

Higher sensitivity can surface more deviations; lower sensitivity focuses on stronger departures. The goal is not simply to find the largest value, but values that are unexpected relative to the time-series pattern.

Outlier and anomaly are not identical

A scatter chart can reveal an isolated numeric observation even without time. Anomaly detection specifically models a time-series pattern and can evaluate other fields to explain factors associated with a marked point.

Some line-chart features cannot be combined with anomaly detection, including certain forecasting scenarios.

Exam clue: unusual point in general = outlier analysis; unexpected time-series point plus expected range = anomaly detection.