MLA-C01 Question 75
Single answerYou are tasked with building a machine learning model for a customer to classify images into different categories. You need a high-quality labeled dataset to train the model. While evaluating different AWS services, you want to ensure that the labeling process is efficient, scalable, and tailored to your specific requirements. Which AWS service should you use to achieve this?
- A
Amazon SageMaker Ground Truth
- B
Amazon Rekognition
- C
AWS Data Pipeline
- D
Amazon Mechanical Turk
Show answer and explanation
Correct answer: A
Explanation
Amazon SageMaker Ground Truth is the best choice for creating high-quality labeled datasets. It provides a scalable and efficient way to label data using either automated labeling techniques, human annotators, or a combination of both. This makes it ideal for use cases like training an image classification model, while other services like Rekognition and Data Pipeline are not designed for this specific purpose.
- A. Correct.
Amazon SageMaker Ground Truth is the appropriate service for creating high-quality labeled datasets. It offers features like built-in workflows, automated labeling to reduce costs, and integration with human labelers.
- B. Incorrect.
Amazon Rekognition is primarily a pre-trained AI service for image and video analysis, such as detecting objects or faces. It does not create custom labeled datasets.
- C. Incorrect.
AWS Data Pipeline is used for data workflow orchestration and movement, not for labeling datasets.
- D. Incorrect.
Amazon Mechanical Turk can be used for human labeling tasks, but it does not provide the built-in workflows, automation, or integration that SageMaker Ground Truth offers specifically for machine learning datasets.