Your question is Choosing Model Evaluation Techniques. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
You are discussing how you evaluate machine learning models before and after deployment. The interviewer wants to hear how you choose metrics, validate performance, and avoid relying on a single number.
What techniques do you use for model evaluation?