

AF
You have trained a model and it looks strong on your holdout set, but the team wants to know whether those results are trustworthy. You need to explain how you would validate the model before it is approved for use.
How do you validate machine learning models?