You must have Administrative permissions to score a model.
Select Insights > Outliers.
After you build a model, you can start a scoring process that evaluates incoming events against the model. The process assigns a score that indicates the degree to which the incoming data varies from typical behavior. By default, Outlier Analytics selects the current date as the scoring start date.
You can only score one model at a time, but you can build another model while a different model is being scored.
To start the scoring process:
Select Configuration > Outlier.
From the Available Models table, select the model that you want to score.
The model must be in Build Complete status before you can score it.
Select Score.
Select the date for which you want to start the scoring process, then click Start.
Because of assumptions about the hours and days that comprise a model, do not use a model that you built with Daylight Savings Time data to score non-Daylight Savings Time data. Conversely, do not use a model that you built with non-Daylight Savings Time data to score Daylight Savings Time data.
(Conditional) To pause scoring because of performance or ingestion issues, select Pause.
If you selected a date in the past to start the scoring process, the scoring job runs frequently to catch up to the current date. To allow any running scoring jobs to complete, wait 15 minutes before performing any other action such as deleting a model or resetting scoring.
(Conditional) To resume the scoring process from the point at which you paused it, select Resume.
Alternatively, to restart the scoring process, select Reset.
To view the scored data when scoring completes, select Insights > Outliers.