To most accurately notify you only of real issues and to learn about specific conditions in your business, you can provide feedback to our model on whether monitor anomalies are expected. This helps our model improve and adjust to your business circumstances over time.
If an anomaly is marked as expected, it will be included in the training data set next time it’s updated. This means the model adjusts based on this data point and will be less likely to flag a similar data point as an anomaly.
If an anomaly is not marked as expected (default), it will be excluded in the training data set next time it’s updated. This means the model will flag similar data points as an anomaly again.
Marking an anomaly as expected
Click a monitor and select
Mark as expected to flag an anomaly as expected. The monitor will be updated immediately and highlight the anomaly in green, indicating that it’s expected.
Marking an anomaly
Anomalies are by default marked as anomalies, but you can revert an anomaly you’ve
Marked as expected by clicking
Mark as anomaly