MLA-C01 Question 394
Select 3You are building a machine learning model to predict customer churn for an e-commerce platform. You want to preprocess and normalize the data before training the model using Amazon SageMaker. Which of the following steps should you take to ensure the data is properly prepared for the training process?
- A
Use Amazon SageMaker Processing Jobs to handle data preprocessing workflows, such as scaling and normalization.
- B
Normalize numerical features to ensure they follow a Gaussian distribution, as many machine learning algorithms assume normalized input data.
- C
Manually write a Python script to preprocess the data locally and upload the processed data to an S3 bucket.
- D
Use Amazon SageMaker Feature Store to store preprocessed features and ensure consistent feature engineering across training and inference.
- E
Rely on the built-in algorithms in Amazon SageMaker to automatically preprocess the data without additional steps.
Show answer and explanation
Correct answers: A, B, D
Explanation
Proper data preprocessing is crucial for building effective machine learning models. Amazon SageMaker provides scalable and efficient tools like Processing Jobs and Feature Store to manage preprocessing workflows. Normalizing numerical features is a standard practice to ensure data is compatible with machine learning models. While SageMaker's built-in algorithms can handle some basic preprocessing, custom workflows are often necessary to meet specific requirements.
- A. Correct.
Correct: Amazon SageMaker Processing Jobs provide an efficient and scalable way to preprocess data, such as scaling and normalization, without requiring manual scripts.
- B. Correct.
Correct: Normalizing numerical features is a common preprocessing step to ensure compatibility with many machine learning algorithms, especially those sensitive to feature scaling.
- C. Incorrect.
Incorrect: While preprocessing locally is possible, it is not the recommended method for scalable and repeatable workflows in Amazon SageMaker.
- D. Correct.
Correct: Using Amazon SageMaker Feature Store helps maintain consistency in feature engineering and ensures the same preprocessing logic is used during both training and inference.
- E. Incorrect.
Incorrect: Although SageMaker's built-in algorithms handle some preprocessing, they do not eliminate the need for explicit steps such as normalization or custom preprocessing pipelines.