MLS-C01 Question 300
Single answerA company wants to build a real-time fraud detection system for its online payment platform. They need to process streaming transaction data and use a pre-trained machine learning model to classify transactions as fraudulent or non-fraudulent. Which AWS service is the most suitable for deploying the fraud detection model in this scenario?
- A
Amazon SageMaker Hosting Services
- B
AWS Lambda
- C
Amazon Kinesis Data Analytics
- D
Amazon Rekognition
Show answer and explanation
Correct answer: A
Explanation
Amazon SageMaker Hosting Services is the best choice for deploying a pre-trained ML model in this scenario because it provides a scalable and managed environment for real-time inference. It integrates seamlessly with Amazon Kinesis for processing streaming data, making it ideal for fraud detection use cases.
- A. Correct.
Amazon SageMaker Hosting Services is the most suitable option for deploying pre-trained machine learning models in real-time scenarios. It provides managed endpoints for hosting models, supports high throughput, and integrates well with other AWS services like Amazon Kinesis for real-time data processing.
- B. Incorrect.
AWS Lambda can execute code in response to events, but it is not optimized for hosting machine learning models that require high throughput or real-time inference. Lambda has resource limitations that make it less ideal for deploying models for continuous, real-time predictions.
- C. Incorrect.
Amazon Kinesis Data Analytics is designed for real-time data stream processing and analytics but does not directly support deploying machine learning models. It can be used in conjunction with other services like SageMaker for preprocessing data.
- D. Incorrect.
Amazon Rekognition is a specialized service for image and video analysis. It is not suitable for deploying machine learning models for fraud detection in transactional data streams.