What You'll Be Tested On

Ingesting and transforming structured, semi-structured, and unstructured datasets
Cleaning, normalizing, encoding, labeling, and validating data before training
Managing datasets in S3 and cataloging them with AWS Glue and Athena
Creating reusable features and preventing data leakage
Checking data quality, class imbalance, outliers, and schema drift

Key AWS Services in This Domain

💡

Start every data scenario by identifying the source, format, quality issue, and target ML workflow.

💡

Use Glue and Athena for discovery and transformation when the data is already in S3.

💡

Feature leakage is a common trap: do not use information that would not exist at prediction time.

💡

Know when an online feature store is needed for low-latency inference and when the offline store supports training.

Test your knowledge of Data Preparation for Machine Learning with practice questions from our MLA-C01 question bank.