Lesson 3 Quiz — Lifecycle, MLOps & Metrics (10 easy + 10 hard)

What does 'MLOps' stand for?

Which is NOT one of the four confusion-matrix outcomes?

Which metric answers 'of everything flagged as fraud, how many were actually fraud'?

Which metric answers 'of all actual fraud cases, how many did I catch'?

True or false: a model can have excellent technical metrics and still be a business failure.

Which deployment method means AWS manages the infrastructure and you just send/receive requests?

What is 'model drift'?

Which pipeline stage typically comes right after data collection?

Which is a business metric rather than a model metric?

Which AWS console page did live verification show has been retired?

A fraud model has high recall but low precision. What does this most likely mean?

Which of the following are TRUE about accuracy as a metric?

Starting right after 'Build & Train', order the remaining MLOps loop stages.

Match each MLOps concept to its definition.

A startup wants a general-purpose LLM quickly, no training from scratch, and wants AWS to handle all infrastructure/scaling. Best combination of (1) FM source, (2) deployment method?

A cancer-screening model missing a real case (false negative) is far more costly than a false alarm. Which metric should be prioritized?

A fraud model has excellent precision/recall but costs $2/prediction while preventing only $0.50 of loss per prediction. Business conclusion?

Order these five ML pipeline stages correctly.

Match each AWS tool to its primary purpose.

Based on live console verification, which statement is accurate about accessing Bedrock foundation models today?