MLA-C01 Question 11
Select 3You are designing a machine learning pipeline to process and classify customer reviews for a retail website. The reviews are stored in an S3 bucket, and you are using Amazon SageMaker to train a text classification model. You want to ensure that your pipeline is efficient and scalable. Which combination of actions should you take?
- A
Use Amazon Comprehend to preprocess the reviews before training the model in SageMaker.
- B
Use an AWS Glue job to extract, transform, and load the data from the S3 bucket into a SageMaker notebook instance.
- C
Use SageMaker Processing Jobs to clean and preprocess the data within the pipeline.
- D
Enable Amazon SageMaker Automatic Model Tuning (Hyperparameter Tuning) to optimize the model performance.
- E
Use an EC2 instance with a custom script to preprocess the data and upload the results back to the S3 bucket.
Show answer and explanation
Correct answers: A, C, D
Explanation
To design an efficient and scalable machine learning pipeline, it is important to utilize AWS managed services that integrate well with each other. Amazon Comprehend allows for advanced text preprocessing, SageMaker Processing Jobs provide a scalable way to clean and preprocess data, and SageMaker Automatic Model Tuning helps optimize model performance. Using AWS Glue or EC2 for this scenario is either overly complex or inefficient compared to the other options.
- A. Correct.
Correct: Amazon Comprehend can be used to preprocess text data, such as extracting key phrases or performing sentiment analysis, before training the model in SageMaker.
- B. Incorrect.
Incorrect: AWS Glue is primarily used for ETL operations and is not the most efficient tool for preprocessing data directly for SageMaker. Using Glue in this scenario adds unnecessary complexity.
- C. Correct.
Correct: SageMaker Processing Jobs are designed for scalable and efficient data preprocessing and can be seamlessly integrated into the machine learning pipeline.
- D. Correct.
Correct: Enabling SageMaker Automatic Model Tuning can help optimize hyperparameters, leading to better model performance.
- E. Incorrect.
Incorrect: While using an EC2 instance with a custom script is possible, it is not scalable and does not leverage SageMaker's managed services, making it less efficient for this use case.