Exclude a class from prediction
-
- UpdatedJan 30, 2025
- 2 minutes to read
- Yokohama
- AI Experiences
Exclude a class from prediction if its precision or coverage aren't satisfactory. Excluding a class prevents the classification model from predicting a particular output field value.
Before you begin
- Train the solution definition whose output field values you want to exclude.
- Role required: admin or ml_admin
About this task
As part of the testing and refinement of your classification solution, you can try excluding an output class to check the effect on the model's results.
Excluding a class from prediction is temporary. The class is restored the next time you train your solution, as long as the solution definition remains the same. If your precision or coverage targets are still not met, consider deactivating the solution or changing the solution definition.
Typically, you exclude a class from prediction if you want a person to manually set the excluded class value. For example, exclude a particular incident category when the solution doesn't offer sufficient precision or coverage, or because the class triggers other business logic that requires review or approval.
Procedure
Result
What to do next