MLS-C01 Question 22
Select 1A retail company is using AWS to build a machine learning pipeline for processing customer transaction data. They need to process large volumes of historical transaction data stored in Amazon S3 to train machine learning models. Additionally, they want to monitor real-time transactions for potential fraudulent activity. Which combination of data job styles and job types should they implement?
- A
Batch processing for historical transaction data and streaming processing for real-time transactions
- B
Streaming processing for historical transaction data and batch processing for real-time transactions
- C
Batch processing for both historical and real-time transaction data
- D
Streaming processing for both historical and real-time transaction data
- E
Batch processing for historical transaction data only, without streaming for real-time transactions
Show answer and explanation
Correct answer: A
Explanation
In this scenario, the company has distinct use cases: processing historical data and monitoring real-time transactions. Batch processing is ideal for historical data as it allows the system to process large datasets in an efficient manner. Streaming processing is essential for real-time data to detect and respond to events like fraud as they occur. Therefore, the correct combination is batch processing for historical data and streaming processing for real-time transactions.
- A. Correct.
Correct: Batch processing is appropriate for processing large volumes of historical data, while streaming processing is ideal for monitoring real-time transactions for fraud detection.
- B. Incorrect.
Incorrect: Streaming processing is not suitable for historical transaction data as it is designed for continuous data flows, not large-scale static datasets.
- C. Incorrect.
Incorrect: Real-time transactions require streaming processing to detect fraud promptly, making this combination inadequate.
- D. Incorrect.
Incorrect: Historical data is better suited for batch processing, and streaming processing alone cannot handle large-scale historical datasets effectively.
- E. Incorrect.
Incorrect: Ignoring real-time transactions means the system cannot monitor for fraud in real-time, which does not meet the requirements.