C
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?
