An organization wants to implement continuous model monitoring for a production AI system. Which combination of metrics provides the MOST comprehensive coverage of model health?
Select an answer to reveal the explanation.
Short Explanation
Here's the deal — you need data drift, prediction drift, and downstream business KPI tracking because they cover input changes, output changes, and real business impact. Training accuracy and model file size don't tell you production health.
Full Explanation
Data drift, prediction drift, and downstream business KPI tracking is correct because comprehensive model monitoring requires tracking data drift (changes in input distribution), prediction drift (changes in output distribution), and downstream business KPIs (whether the model is still achieving its intended business purpose). Training accuracy and model file size focus on static training metrics, not production health. Infrastructure uptime and API response latency measure infrastructure, not model quality. Developer commit frequency and code coverage percentage measure development process, not deployed model performance.