AWS Certified Machine Learning Engineer - Associate Exam
65-question written exam with multiple-choice, multiple-response, ordering, and matching question types.
- Type
- Written
- Delivery
- Both
- Duration
- 130 min
- Questions
- 65
Passing score: 720 Scaled score on a 100 to 1,000 scale
Exam sections
Domain 1: Data Preparation for Machine Learning (ML)
Focuses on ingesting, storing, and transforming data for machine learning. Key tasks include merging data from multiple sources using AWS Glue, performing feature engineering like scaling and encoding, and ensuring data integrity by mitigating bias and validating quality with SageMaker Clarify and DataBrew.
Question notes
Includes multiple choice, multiple response, ordering, and matching questions.
Preparation tips
Focus on AWS data ingestion services (Kinesis, Glue) and feature engineering tools like SageMaker Data Wrangler and Feature Store.
Domain 2: ML Model Development
Covers choosing modeling approaches, training, and refining models. It includes using SageMaker AI built-in algorithms, fine-tuning foundation models in Bedrock, and analyzing performance with metrics like F1 score and RMSE. Key skills involve preventing overfitting and managing model versions in the SageMaker Model Registry.
Question notes
Includes multiple choice, multiple response, ordering, and matching questions.
Preparation tips
Master SageMaker built-in algorithms, hyperparameter tuning (AMT), and model evaluation metrics for both regression and classification.
Domain 3: Deployment and Orchestration of ML Workflows
Involves selecting deployment infrastructure and orchestrating ML workflows. It focuses on serving models in real time or batches using SageMaker endpoints, implementing CI/CD pipelines with AWS CodePipeline, and automating resource provisioning using Infrastructure as Code tools like AWS CDK and CloudFormation.
Question notes
Includes multiple choice, multiple response, ordering, and matching questions.
Preparation tips
Understand different SageMaker endpoint types (Serverless, Asynchronous, Real-time) and CI/CD tools like CodeBuild and SageMaker Pipelines.
Domain 4: ML Solution Monitoring, Maintenance, and Security
Focuses on monitoring model inference, infrastructure, and security. Tasks include detecting data drift with SageMaker Model Monitor, optimizing costs using AWS Compute Optimizer and Savings Plans, and securing resources through IAM policies, VPC configurations, and encryption to ensure compliance and least privilege access.
Question notes
Includes multiple choice, multiple response, ordering, and matching questions.
Preparation tips
Study SageMaker Model Monitor for drift detection, AWS Cost Explorer for cost management, and IAM/VPC security best practices for ML.