AWS Certified Machine Learning – Specialty — Question 341

A company has 2,000 retail stores. The company needs to develop a new model to predict demand based on holidays and weather conditions. The model must predict demand in each geographic area where the retail stores are located.

Before deploying the newly developed model, the company wants to test the model for 2 to 3 days. The model needs to be robust enough to adapt to supply chain and retail store requirements.

Which combination of steps should the company take to meet these requirements with the LEAST operational overhead? (Choose two.)

Answer options

• A. Develop the model by using the Amazon Forecast Prophet model.
• B. Develop the model by using the Amazon Forecast holidays featurization and weather index.
• C. Deploy the model by using a canary strategy that uses Amazon SageMaker and AWS Step Functions.
• D. Deploy the model by using an A/B testing strategy that uses Amazon SageMaker Pipelines.
• E. Deploy the model by using an A/B testing strategy that uses Amazon SageMaker and AWS Step Functions.

Correct answer: B, C

Explanation

Amazon Forecast provides built-in support for holidays featurization and a weather index, which directly addresses the demand forecasting requirements with minimal custom development. To test the model safely over 2 to 3 days with low operational overhead, a canary deployment using Amazon SageMaker and AWS Step Functions allows for gradual traffic shifting and automated rollback if issues arise. A/B testing is less appropriate here as the goal is to validate a single new model's robustness before full release rather than comparing two active variants.