• Make sure you have admin user privileges to configure the training settings.
  • If you're accessing PingIntelligence through PingOne platform, make sure that the PingOne API Intelligence service is activated. For more information, see Getting started with PingOne API Intelligence.
The following steps provide an overview of the different training parameters that can be configured through the Dashboard. It is recommended that you review the variables and configure the best values for your environment.
  • Click Settings -> Training Settings as shown in the following screenshot. By default you will reach the Global Configuration page.

    Screenshot for training settings page

    Note: You need admin user privileges to update the Training settings.
  • You can configure the following two settings in the Global Configuration page.
    • TRAINING PERIOD- The number of hours to train the AI model before it moves to attack detection mode. It is recommended that you configure the training for at least a week's direction, in a production environment.
    • TRAINING UPDATE INTERVAL- The time interval at which continuous learning model thresholds are updated in ABS AI engine.
    These variables are specified in hours with an allowable range of 1 to 10000. Click Save on the bottom-left to reflect the changes.

  • You can reset the training of an API or multiple APIs from the Reset Training page. To reset the training, select the APIs in RESET TRAINING and click Go.
    Note: If there are any pending jobs in the ABS AI engine, the reset will fail with an error notification. You can re-run resetting after a few minutes. If the reset fails multiple times, follow the steps explained in Resetting trained APIs in ABS to manually reset the API.

    Screen shot for Reset training