AWS Certified Machine Learning Engineer Associate · Free Practice Question Medium
Question 8
A pharmaceutical company is using machine learning to analyze clinical trial data. They need to ingest raw data from multiple sources, such as CSV files and relational databases, and clean and join this data. Afterward, the data should be used to create a set of features, including patient age, treatment history, and genetic markers. The company also wants to maintain a version-controlled feature repository that can be reused across different machine learning models.
Which AWS services should the company use to preprocess the data, create features, and maintain a version-controlled feature repository?
-
A
Use AWS Glue for data cleaning, store features in Amazon DynamoDB, and use SageMaker to train models.
-
B
Use Amazon Athena to preprocess the data, store the features in Amazon S3, and use SageMaker Autopilot to automatically train models.
-
C
Use Amazon QuickSight for data preprocessing, SageMaker Feature Store to store features, and SageMaker JumpStart to deploy a pre-built model.
-
D
Use SageMaker Data Wrangler for data preprocessing, SageMaker Feature Store to store and manage the features, and SageMaker Notebooks to train the models.
Reveal correct answer
Correct answer: D
A.
AWS Glue is primarily used for ETL (extract, transform, load) processes, but it does not provide the version-controlled feature storage needed for machine learning. DynamoDB is not optimal for storing machine learning features compared to SageMaker Feature Store.
B.
Amazon Athena is not designed for complex data cleaning and feature engineering. Amazon S3 does not offer version-controlled feature storage like SageMaker Feature Store does.
C.
Amazon QuickSight is a data visualization tool and is not intended for preprocessing or feature engineering. SageMaker Data Wrangler would be a better fit for this task.
D.
SageMaker Data Wrangler allows for seamless data cleaning and feature engineering. SageMaker Feature Store provides version control and centralized management of features, making it easy to reuse features across multiple models. SageMaker Notebooks offer an integrated environment for training and evaluating models.
Discussion
Think the marked answer is wrong, or have a better explanation? Share it below — comments appear after review.
