Model validation and testing
Checking that a model is fit for its intended use before and after deployment.
Validation covers conceptual soundness, data quality, performance on held-out data, fairness, robustness, explainability and implementation, comparing results with the requirements set for the use case.
On the exam
Validation tests against requirements set before development. Performance shown only on training data isn’t validation.
Related terms
- Independent model validationValidation by people who didn’t build the model.
- Accuracy, precision, recall and F1 scoreMeasures of how well a classification model performs.
- Robustness and stress testingChecking that a model holds up under unusual, noisy or extreme inputs.
AAIA Test yourself on this with timed AAIA practice questions in the AAIA Exam Simulation Hub, or read the free AAIA exam guide.
Test yourself with today’s question of the day.
