Skip to main content
Documentation
Documentation
Adaptive Planning
Last Updated: 2023-08-04
Detect Anomalies in Sheets

Detect Anomalies in Sheets

Anomaly Detection
 uses a machine learning algorithm to predict values in a prediction version. In any other plan data, the machine compares the data to those predictions. When values fall outside of the predicted range of the algorithm, standard sheets highlight the cells with a purple border.
When you use anomaly detection to identify accounting anomalies, you can discover hidden trends and outliers in your data.
Anomaly detection works for:
  • General ledger and custom account data.
  • Periodic account types only.
  • Plan versions, but not the prediction version. The prediction version holds the predictions for the algorithm.
  • Standard sheets without dimensions.
  • Web sheets, not Excel Interface for Planning sheets.
  2. From the sheet toolbar, select
    Show Anomalies
    .
    We display purple borders around cells with anomalies. You can select
    Show Anomalies
     again to hide the purple anomaly borders.
    After you save or refresh the sheet, you must select
    Show Anomalies
     again to reveal anomalies.
  3. (Optional) Hover over any cell with an anomaly to review information about how the anomaly was detected:
    Field
    Description
    Predicted value
    Displays what the algorithm thinks the value should be.
    Lower bound
    Upper bound
    Displays the predicted range of acceptable values from the algorithm.
  4. (Optional) To correct a cell so that it no longer displays as an anomaly, right-click the cell and select
    Unmark Anomaly
    .
    We remove the purple border when you save the sheet.