DJames617 MLA-C01 / Domain 2 / 26%
ML Model Development
Model selection, training, tuning, evaluation, and refinement.
Official Task Statements
| Task | What to prove |
|---|---|
| MLA-2.1 | Choose a modeling approach. |
| MLA-2.2 | Train and refine models. |
| MLA-2.3 | Analyze model performance. |
Concepts You Need to Understand
- Model choice, built-in algorithms, custom containers, training jobs, hyperparameter tuning, evaluation metrics, bias, and explainability.
AWS services involved
- SageMaker AI
- SageMaker Training
- SageMaker Experiments
- SageMaker Clarify
- SageMaker Debugger
Important configurations
- Training instance type.
- Input channels.
- Tuning ranges.
- Evaluation metric.
- Experiment tracking.
Exam Decision Patterns
Least operational overhead
Prefer managed and serverless services when they satisfy the requirement. Exceptions appear when the scenario needs host control, unsupported runtimes, specialized network behavior, or exact migration compatibility.
Highly available
Identify the failure boundary. One instance is not HA. Multiple instances in one AZ help capacity but not AZ failure. Multi-AZ handles regional AZ faults. Multi-Region handles regional events but adds complexity and cost.
Durable
Durability is about preserving data. Use replication, versioning, backups, point-in-time recovery, and tested restore plans. A durable backup does not guarantee a low RTO.
Decouple the application
Use SQS for buffering work, SNS for fanout, EventBridge for event routing, and Step Functions for visible workflow state. Add retries, DLQs, and idempotent consumers.
Least privilege
Prefer roles and temporary credentials, scope actions/resources/conditions, watch explicit denies, and remember that resource policies may also be required.
Most cost-effective
Read usage pattern, duration, access frequency, scaling behavior, data transfer, and operations. Cheapest unit price is not always lowest total cost.
Lowest latency
Move content or compute closer to users, cache aggressively, choose the right database access pattern, and avoid unnecessary cross-Region or NAT paths.
Private connectivity
Use private subnets, VPC endpoints, PrivateLink, VPN, Direct Connect, Transit Gateway, and tight DNS/routing design instead of public exposure.
Minimum downtime
Separate deployment downtime, failure recovery, and data restore time. Use blue/green, canary, Multi-AZ, replication, and tested rollback where appropriate.
Automatic remediation
Pair a reliable signal with EventBridge or CloudWatch, a scoped Systems Manager Automation or Lambda action, and a validation step.
Common Mistakes
- Optimizing accuracy when precision, recall, latency, or cost is the real requirement.
- Ignoring class imbalance.
Example Architecture
Hands-On Activity
Choose evaluation metrics for fraud, image labeling, and recommendation scenarios.
For an AWS-account lab, use one of the linked mini labs and keep cleanup steps visible before you start.
Task-by-Task Study Notes
MLA-2.1 - Choose a modeling approach.
Read this task as a decision problem: identify the workload requirement, the control or service family involved, and the tradeoff AWS is testing. The validated local corpus connects this objective to 6 official AWS sources and identifies these study anchors:
- Choose a modeling approach.
- ML Model Development
- SageMaker AI
- SageMaker Training
- SageMaker Experiments
- SageMaker Clarify
- SageMaker Debugger
How to apply the material
- Translate the wording into requirements: security, operations, cost, availability, latency, governance, or data behavior.
- Choose the service or configuration that directly satisfies those requirements with the least unnecessary complexity.
- Reject options that are technically possible but miss the domain goal or increase risk without a requirement.
Explain before memorizing
Know this distinction: explain why the selected approach fits the requirement, what it does not provide, and which customer-managed control remains. Then test the explanation against a changed constraint: a different failure boundary, traffic pattern, data sensitivity, latency target, or operating-cost limit.
Exam habit: when two answers seem technically possible, prefer the one that matches the stated outcome and shared-responsibility boundary. Do not assume that a managed service removes identity, data-protection, configuration, monitoring, recovery, or cost responsibilities.
Mastery check before Arcade practice
- I can define the central terms and explain what problem the objective is solving.
- I can select the best answer from a realistic scenario without relying on a product name alone.
- I can explain why the closest distractor is wrong when one requirement changes.
- I can identify the AWS-managed boundary and the customer-managed control that remains.
- I can predict the main availability, security, scaling, operations, or cost consequence of the choice.
Use the Arcade after you can explain all five checks aloud or in writing. If an answer is correct only because it looks familiar, return to the source anchors and compare the service purpose, constraints, and tradeoffs again.
Practice MLA-2.1 style questions in this domain
Official AWS references for this objective
- AWS docs / aws-certification/latest/machine-learning-engineer-associate-01/machine-learning-engineer-associate-01.html
- AWS docs / sagemaker/latest/dg/whatis.html
- AWS docs / sagemaker/latest/dg/model-monitor.html
- AWS docs / AmazonS3/latest/userguide/Welcome.html
- AWS docs / IAM/latest/UserGuide/introduction.html
Sources are from the validated local corpus; retrieved 2026-08-14. Retrieval metadata and hashes are retained in the corpus manifest.
MLA-2.2 - Train and refine models.
Read this task as a decision problem: identify the workload requirement, the control or service family involved, and the tradeoff AWS is testing. The validated local corpus connects this objective to 6 official AWS sources and identifies these study anchors:
- Train and refine models.
- ML Model Development
- SageMaker AI
- SageMaker Training
- SageMaker Experiments
- SageMaker Clarify
- SageMaker Debugger
How to apply the material
- Translate the wording into requirements: security, operations, cost, availability, latency, governance, or data behavior.
- Choose the service or configuration that directly satisfies those requirements with the least unnecessary complexity.
- Reject options that are technically possible but miss the domain goal or increase risk without a requirement.
Explain before memorizing
Know this distinction: explain why the selected approach fits the requirement, what it does not provide, and which customer-managed control remains. Then test the explanation against a changed constraint: a different failure boundary, traffic pattern, data sensitivity, latency target, or operating-cost limit.
Exam habit: when two answers seem technically possible, prefer the one that matches the stated outcome and shared-responsibility boundary. Do not assume that a managed service removes identity, data-protection, configuration, monitoring, recovery, or cost responsibilities.
Mastery check before Arcade practice
- I can define the central terms and explain what problem the objective is solving.
- I can select the best answer from a realistic scenario without relying on a product name alone.
- I can explain why the closest distractor is wrong when one requirement changes.
- I can identify the AWS-managed boundary and the customer-managed control that remains.
- I can predict the main availability, security, scaling, operations, or cost consequence of the choice.
Use the Arcade after you can explain all five checks aloud or in writing. If an answer is correct only because it looks familiar, return to the source anchors and compare the service purpose, constraints, and tradeoffs again.
Practice MLA-2.2 style questions in this domain
Official AWS references for this objective
- AWS docs / aws-certification/latest/machine-learning-engineer-associate-01/machine-learning-engineer-associate-01.html
- AWS docs / sagemaker/latest/dg/whatis.html
- AWS docs / sagemaker/latest/dg/model-monitor.html
- AWS docs / AmazonS3/latest/userguide/Welcome.html
- AWS docs / IAM/latest/UserGuide/introduction.html
Sources are from the validated local corpus; retrieved 2026-08-14. Retrieval metadata and hashes are retained in the corpus manifest.
MLA-2.3 - Analyze model performance.
Read this task as a decision problem: identify the workload requirement, the control or service family involved, and the tradeoff AWS is testing. The validated local corpus connects this objective to 6 official AWS sources and identifies these study anchors:
- Analyze model performance.
- ML Model Development
- SageMaker AI
- SageMaker Training
- SageMaker Experiments
- SageMaker Clarify
- SageMaker Debugger
How to apply the material
- Translate the wording into requirements: security, operations, cost, availability, latency, governance, or data behavior.
- Choose the service or configuration that directly satisfies those requirements with the least unnecessary complexity.
- Reject options that are technically possible but miss the domain goal or increase risk without a requirement.
Explain before memorizing
Know this distinction: explain why the selected approach fits the requirement, what it does not provide, and which customer-managed control remains. Then test the explanation against a changed constraint: a different failure boundary, traffic pattern, data sensitivity, latency target, or operating-cost limit.
Exam habit: when two answers seem technically possible, prefer the one that matches the stated outcome and shared-responsibility boundary. Do not assume that a managed service removes identity, data-protection, configuration, monitoring, recovery, or cost responsibilities.
Mastery check before Arcade practice
- I can define the central terms and explain what problem the objective is solving.
- I can select the best answer from a realistic scenario without relying on a product name alone.
- I can explain why the closest distractor is wrong when one requirement changes.
- I can identify the AWS-managed boundary and the customer-managed control that remains.
- I can predict the main availability, security, scaling, operations, or cost consequence of the choice.
Use the Arcade after you can explain all five checks aloud or in writing. If an answer is correct only because it looks familiar, return to the source anchors and compare the service purpose, constraints, and tradeoffs again.
Practice MLA-2.3 style questions in this domain
Official AWS references for this objective
- AWS docs / aws-certification/latest/machine-learning-engineer-associate-01/machine-learning-engineer-associate-01.html
- AWS docs / sagemaker/latest/dg/whatis.html
- AWS docs / sagemaker/latest/dg/model-monitor.html
- AWS docs / AmazonS3/latest/userguide/Welcome.html
- AWS docs / IAM/latest/UserGuide/introduction.html
Sources are from the validated local corpus; retrieved 2026-08-14. Retrieval metadata and hashes are retained in the corpus manifest.