Your question is Metrics for ML Project Success. 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).
What metrics would you use to assess the success of a machine learning project?
Explain how your choices would vary by task type, class balance, error costs, and business objective. Address offline validation, comparison with a baseline, and monitoring after deployment.
