Exit 15 / 40

Question 15

Content Domain 3: Modeling

A machine learning team has already measured model quality on a held-out validation dataset, but now wants to determine whether the new model improves actual user behavior after deployment. Which evaluation approach should the team use next?