Skip to content

Click in-app to access the full platform documentation for your version of DataRobot.

Multiclass support in No-Code AI Apps

Availability information

Support for multiclass projects in No-Code AI Apps is off by default. Contact your DataRobot representative or administrator for information on enabling the feature.

Feature flag: Enable Application Builder Multiclass Support

In addition to regression and binary classification deployments, No-Code AI Apps now support multiclass classification deployments across all three template types—Predictor, Optimizer, and What-If. This gives users the ability to create applications that solve a broader range of business problems, for example, deciding which customers should be contacted for specific promotions and which customers should not be contacted for a promotion. Because each customer is unique, offering the same promotions to every customer may cause them to churn. Multiclass applications can help accurately predict the most relevant promotions for each customer.

To create a multiclass classification application:

  1. Create an application from your multiclass deployment.
  2. In Build mode, select the All Rows widget and click Manage.
  3. Use the orange arrows to add the predicted class field to the Selected features column and click Save.

    The All Rows widget updates to display the predicted class field.

  4. If you create an Optimizer application, you need to select a target for the optimizer from the predicted class values. To do so, select One Row from the Active Page dropdown and select the What-if and Optimizer widget. Then in the left panel, select Properties and use the dropdown to select your target. Note that Enable scenario optimizer must be toggled on.

  5. In Consume mode, upload a batch prediction and view the prediction details, or make a single record prediction.

  6. Under Prediction Information, select the predicted class you want to view using the dropdown. The prediction details page updates based on your selection.

Note

For unlimited multiclass projects with more than 1000 classes, by default, DataRobot keeps the top 999 most frequent classes and aggregates the remainder into a single "other" bucket. You can then select the deployment you'd like to use when creating the application. To prevent this, you can modify the aggregation settings.


Updated June 24, 2022
Back to top