AWS Certified Machine Learning Engineer – Associate (MLA-C01) — Question 200

A healthcare company uses an Amazon SageMaker AI endpoint to host a model that predicts patient readmission risk to hospitals. The company wants to predict patient readmissions with high accuracy and is willing to tolerate false positives. The current model performance has degraded over the previous year.

The company trains and deploys a new model as a shadow variant for testing on live traffic from hospitals. The company monitors the performance of the new model for a month. During the month of testing, the shadow variant has a higher recall than the existing model but has a lower precision.

What should the company do next?

Answer options

• A. Promote the shadow variant to full production.
• B. Extend the shadow testing period to capture more data. Monitor the new model to determine whether precision improves.
• C. Use a blue/green deployment strategy to allocate a small percentage of traffic to the shadow variant to reduce model errors.
• D. Disable the shadow variant and roll back to the main variant.

Correct answer: A

Explanation

Promoting the shadow variant to full production is appropriate since it has demonstrated higher recall, which is crucial for reducing patient readmissions, even though precision is lower. Extending the shadow testing period may help improve precision, but it delays the benefits of the new model. A blue/green deployment could manage traffic but is unnecessary if the shadow variant is already performing better in recall. Disabling the shadow variant would ignore the improvements it has shown.