MLS-C01 Question 178
Single answerA retail company wants to predict customer churn based on historical transaction data. The machine learning model needs to process large volumes of data and generate predictions for all customers at the end of every month. Which approach should the company use to implement the prediction workflow?
- A
Batch inference with Amazon SageMaker Batch Transform
- B
Real-time inference with Amazon SageMaker Endpoint
- C
Event-driven inference with AWS Lambda
- D
Stream processing with Amazon Kinesis Data Analytics
Show answer and explanation
Correct answer: A
Explanation
Batch processing is ideal when predictions are required for a large dataset at a scheduled interval, such as at the end of the month. Amazon SageMaker Batch Transform is specifically designed for batch inference tasks, where a trained model processes datasets in bulk to generate predictions. In this case, real-time or event-driven solutions are not necessary since there is no requirement for immediate responses or streaming data.
- A. Correct.
Batch inference with Amazon SageMaker Batch Transform is the correct approach because the company needs predictions for all customers at the end of each month, which aligns with a batch processing workflow.
- B. Incorrect.
Real-time inference with Amazon SageMaker Endpoint is not suitable because the requirement is not for individual predictions in real-time, but rather for generating predictions in bulk at a scheduled interval.
- C. Incorrect.
Event-driven inference with AWS Lambda is incorrect because there is no mention of an event-driven system or real-time triggers for individual predictions.
- D. Incorrect.
Stream processing with Amazon Kinesis Data Analytics is incorrect because the use case does not involve real-time data streams, but rather historical data processed in batches.