Managing ML models and templates
The Models section provides data on ML models and on the templates created based on the ML models added to Kaspersky MLAD.
ML models and templates can be managed only by the users with administrator privileges.
Models tab
The Models tab shows a table of ML models. You can upload ML models to Kaspersky MLAD or add ML models based on ML model templates.
The following information is displayed for each ML model:
- ID – digital identifier of the ML model.
- Model name – name of the added ML model.
- Creation method – indicator of whether the ML model was imported, created from a template, or created as a result of training on a previously added ML model. If the template used to create the ML model is deleted, the Creation method column shows the Template deleted value. If the ML model used to train the current ML model is deleted, the Creation method column shows the Cloned and trained; original model deleted value.
- State – toggle button allowing you to enable or disable the use of the ML model for the monitored asset. If an ML model is not activated, the State column shows an inactive Unavailable toggle button.
- Action – button for activating the ML model or starting the ML model training process after activation.
Next to each model, there is a vertical menu , allowing you to create a template based on the existing ML model, view detailed information about the model, or delete the ML model.
Templates tab
The Templates tab provides a table of templates that were created based on ML models added to Kaspersky MLAD. The following information is available for each template:
- ID – digital identifier of the ML model template.
- Template name – name of the ML model template.
- Creation method – name of the ML model used to create the template. If the ML model used to create the template is deleted, the Creation method column shows the Original model was deleted value.
- Action – button for creating an ML model based on a template.
You can use created templates to add new ML models to Kaspersky MLAD. After adding an ML model, you can enable its use for the monitored asset. You can also delete ML model templates.