cert
cert-1
cert-2

Pass Amazon AWS Certified Machine Learning Engineer - Associate MLA-C01 Exam in First Attempt Guaranteed!

Get 100% Latest Exam Questions, Accurate & Verified Answers to Pass the Actual Exam!
30 Days Free Updates, Instant Download!

cert-5
cert-6
AWS Certified Machine Learning Engineer - Associate MLA-C01 Exam - Verified By Experts
AWS Certified Machine Learning Engineer - Associate MLA-C01 Premium Bundle
$29.98

AWS Certified Machine Learning Engineer - Associate MLA-C01 Premium Bundle

$79.99
$109.97
  • Premium File 270 Questions & Answers. Last update: Oct 24, 2026
  • Training Course 244 Video Lectures
  • Study Guide 548 Pages
 
$109.97
$79.99
accept 42 downloads in last 7 days
block-screenshots
Purchase Amazon AWS Certified Machine Learning Engineer - Associate MLA-C01 Exam Training Products Individually
Amazon AWS Certified Machine Learning Engineer - Associate MLA-C01 Premium File
Premium File 270 Q&A
$76.99$69.99
Amazon AWS Certified Machine Learning Engineer - Associate MLA-C01 Video Training Course
Training Course 244 Lectures
$21.99 $19.99
Amazon AWS Certified Machine Learning Engineer - Associate MLA-C01 PDF Study Guide
Study Guide 548 Pages
$21.99 $19.99
Amazon AWS Certified Machine Learning Engineer - Associate MLA-C01 Practice Test 9

Question 80

At a production AWS ML data lake supporting customer-facing predictions, the following condition is observed: A rare-defect classifier oversamples minority instances before creating cross-validation folds, causing near copies in both training and holdout. Engineers must avoid changing unrelated source partitions. Which observation most directly identifies the root cause of the reported behavior?

  1. Inspect Spark stage durations, partition size distribution, join key frequencies, and executor spill metrics.
  2. Inspect sampler placement, duplicate neighbor IDs, fold membership, and class-conditioned metric changes.
  3. Measure file-size distribution, object request count, Spark task scheduling overhead, and read throughput.
  4. Inspect feature availability timestamps and whether each label time follows the last feature observation.

Explanation

Correct Answer

B:

Inspect sampler placement, duplicate neighbor IDs, fold membership, and class-conditioned metric changes. This directly tests class imbalance handled inside cross-validation in the reported AWS ML workflow. Oversampling before splitting contaminates evaluation and hides the real error rate on rare classes. It distinguishes the problem before choosing a change. The related targeted response is: Apply resampling only inside each training fold and evaluate untouched validation folds using precision and recall metrics. Reported signal: A rare-defect classifier oversamples minority instances before creating cross-validation folds, causing near copies in both training and holdout. Operational setting: At a production AWS ML data lake supporting customer-facing predictions.

Incorrect Answers

A:

Inspect Spark stage durations, partition size distribution, join key frequencies, and executor spill metrics. This is a possible diagnostic for glue spark join skew on dominant identifier, but it does not directly test the class imbalance handled inside cross-validation failure reported here. Prioritize: Inspect sampler placement, duplicate neighbor IDs, fold membership, and class-conditioned metric changes. The relevant case is: A rare-defect classifier oversamples minority instances before creating cross-validation folds, causing near copies in both training and holdout. Operational setting: At a production AWS ML data lake supporting customer-facing predictions.

C:

Measure file-size distribution, object request count, Spark task scheduling overhead, and read throughput. This is a possible diagnostic for columnar file compaction for small objects, but it does not directly test the class imbalance handled inside cross-validation failure reported here. Prioritize: Inspect sampler placement, duplicate neighbor IDs, fold membership, and class-conditioned metric changes. The relevant case is: A rare-defect classifier oversamples minority instances before creating cross-validation folds, causing near copies in both training and holdout. Operational setting: At a production AWS ML data lake supporting customer-facing predictions.

D:

Inspect feature availability timestamps and whether each label time follows the last feature observation. This is a possible diagnostic for chronological split for forecast evaluation, but it does not directly test the class imbalance handled inside cross-validation failure reported here. Prioritize: Inspect sampler placement, duplicate neighbor IDs, fold membership, and class-conditioned metric changes. The relevant case is: A rare-defect classifier oversamples minority instances before creating cross-validation folds, causing near copies in both training and holdout. Operational setting: At a production AWS ML data lake supporting customer-facing predictions.

Question 81

During production evaluation of a SageMaker AI model candidate, the following condition is observed: A SageMaker PyTorch training container repeatedly runs out of CUDA memory after batch size rises, although the model previously trained. The release is blocked until the underlying model behavior is understood. What targeted change should be implemented after verifying the stated cause?

  1. Compare RMSE with MAE or robust error summaries and choose the metric matching the application's penalty for large misses.
  2. Reduce per-device batch size or use appropriate gradient accumulation or mixed precision after compatibility validation.
  3. Select a licensed, task-appropriate model and validate its performance and resource needs before a controlled deployment.
  4. Adapt the custom training container to SageMaker training job conventions and verify locally against a representative channel layout.

Explanation

Correct Answer

B:

Reduce per-device batch size or use appropriate gradient accumulation or mixed precision after compatibility validation. This is the appropriate controlled remedy for training batch size causes gpu memory exhaustion. Per-step activation memory generally scales with batch size; safe accumulation can preserve an effective larger batch. Verify the cause with: Inspect accelerator memory use, tensor shapes, batch size, gradient accumulation, and failed training logs. Then confirm: Training completes within GPU memory limits and reaches comparable optimizer behavior and accuracy. Reported signal: A SageMaker PyTorch training container repeatedly runs out of CUDA memory after batch size rises, although the model previously trained. Operational setting: During production evaluation of a SageMaker AI model candidate.

Incorrect Answers

A:

Compare RMSE with MAE or robust error summaries and choose the metric matching the application's penalty for large misses. This addresses a different issue involving rmse versus mae outlier sensitivity. It does not correct the stated training batch size causes gpu memory exhaustion mechanism. The targeted response is: Reduce per-device batch size or use appropriate gradient accumulation or mixed precision after compatibility validation. Case evidence: A SageMaker PyTorch training container repeatedly runs out of CUDA memory after batch size rises, although the model previously trained. Operational setting: During production evaluation of a SageMaker AI model candidate.

C:

Select a licensed, task-appropriate model and validate its performance and resource needs before a controlled deployment. This addresses a different issue involving sagemaker jumpstart template provenance. It does not correct the stated training batch size causes gpu memory exhaustion mechanism. The targeted response is: Reduce per-device batch size or use appropriate gradient accumulation or mixed precision after compatibility validation. Case evidence: A SageMaker PyTorch training container repeatedly runs out of CUDA memory after batch size rises, although the model previously trained. Operational setting: During production evaluation of a SageMaker AI model candidate.

D:

Adapt the custom training container to SageMaker training job conventions and verify locally against a representative channel layout. This addresses a different issue involving training image compatibility for byoc. It does not correct the stated training batch size causes gpu memory exhaustion mechanism. The targeted response is: Reduce per-device batch size or use appropriate gradient accumulation or mixed precision after compatibility validation. Case evidence: A SageMaker PyTorch training container repeatedly runs out of CUDA memory after batch size rises, although the model previously trained. Operational setting: During production evaluation of a SageMaker AI model candidate.

Question 82

As a regulated ML workload prepares for an automated regional rollout, the following condition is observed: A real-time recommender fetches cached customer history updated hourly while offline training used a daily snapshot with different aggregation boundaries. The team must demonstrate a repeatable monitoring or security control. Which TWO complementary choices provide the decisive diagnosis AND corrective response? Select TWO.

  1. Use a consistent versioned feature definition and test offline/online values at the same observation cutoff.
  2. Use appropriate SageMaker Clarify model-bias monitoring and investigate supported group differences before deployment changes.
  3. Inspect resource tags, cost allocation tag activation, Cost Explorer grouping, and unallocated resource types.
  4. Inspect online versus offline feature computation code, event-time cutoffs, freshness, and sampled entity values.
  5. Grant only the needed S3 and KMS actions for the approved data prefix and ensure appropriate role trust.

Explanation

Correct Answers

A:

Use a consistent versioned feature definition and test offline/online values at the same observation cutoff. This is the appropriate controlled remedy for training-serving feature skew incident. Mismatched feature windows and update cadence can degrade production predictions even if training accuracy is stable. Verify the cause with: Inspect online versus offline feature computation code, event-time cutoffs, freshness, and sampled entity values. Then confirm: Parity checks match transformations within documented freshness tolerances for test entities. This is one part of a select-two decision; the matching evidence and corrective action are both required. Reported signal: A real-time recommender fetches cached customer history updated hourly while offline training used a daily snapshot with different aggregation boundaries. Operational setting: As a regulated ML workload prepares for an automated regional rollout.

D:

Inspect online versus offline feature computation code, event-time cutoffs, freshness, and sampled entity values. This directly tests training-serving feature skew incident in the reported AWS ML workflow. Mismatched feature windows and update cadence can degrade production predictions even if training accuracy is stable. It distinguishes the problem before choosing a change. The related targeted response is: Use a consistent versioned feature definition and test offline/online values at the same observation cutoff. This is one part of a select-two decision; the matching evidence and corrective action are both required. Reported signal: A real-time recommender fetches cached customer history updated hourly while offline training used a daily snapshot with different aggregation boundaries. Operational setting: As a regulated ML workload prepares for an automated regional rollout.

Incorrect Answers

B:

Use appropriate SageMaker Clarify model-bias monitoring and investigate supported group differences before deployment changes. This selection concerns clarify postdeployment group bias monitoring rather than the required diagnostic-and-remedy pair for training-serving feature skew incident. Use evidence: Inspect online versus offline feature computation code, event-time cutoffs, freshness, and sampled entity values. and the matching correction: Use a consistent versioned feature definition and test offline/online values at the same observation cutoff. Signal: A real-time recommender fetches cached customer history updated hourly while offline training used a daily snapshot with different aggregation boundaries. Operational setting: As a regulated ML workload prepares for an automated regional rollout.

C:

Inspect resource tags, cost allocation tag activation, Cost Explorer grouping, and unallocated resource types. This selection concerns tag-based ml training cost attribution rather than the required diagnostic-and-remedy pair for training-serving feature skew incident. Use evidence: Inspect online versus offline feature computation code, event-time cutoffs, freshness, and sampled entity values. and the matching correction: Use a consistent versioned feature definition and test offline/online values at the same observation cutoff. Signal: A real-time recommender fetches cached customer history updated hourly while offline training used a daily snapshot with different aggregation boundaries. Operational setting: As a regulated ML workload prepares for an automated regional rollout.

E:

Grant only the needed S3 and KMS actions for the approved data prefix and ensure appropriate role trust. This selection concerns sagemaker execution role s3 least privilege rather than the required diagnostic-and-remedy pair for training-serving feature skew incident. Use evidence: Inspect online versus offline feature computation code, event-time cutoffs, freshness, and sampled entity values. and the matching correction: Use a consistent versioned feature definition and test offline/online values at the same observation cutoff. Signal: A real-time recommender fetches cached customer history updated hourly while offline training used a daily snapshot with different aggregation boundaries. Operational setting: As a regulated ML workload prepares for an automated regional rollout.

Question 83

At the model-governance gate before promotion of a new ML artifact, the following condition is observed: A tabular model shows one post-outcome administrative flag as its strongest feature in SageMaker Clarify despite implausibly high offline scores. The team needs a defensible technical decision rather than an untracked workaround. Which engineering response resolves the actual issue without masking the symptoms?

  1. Compare RMSE with MAE or robust error summaries and choose the metric matching the application's penalty for large misses.
  2. Configure SageMaker Automatic Model Tuning with a bounded search space, objective metric, budget, and justified Bayesian or random strategy.
  3. Use a time-series forecasting approach with past context and seasonal features, evaluating all intended future horizons.
  4. Remove label-derived or post-outcome features, rebuild a point-in-time split, and recompute explanations and holdout metrics.

Explanation

Correct Answer

D:

Remove label-derived or post-outcome features, rebuild a point-in-time split, and recompute explanations and holdout metrics. This is the appropriate controlled remedy for clarify feature attribution against target leakage. Attribution can expose suspicious shortcuts; a high importance value is not proof a feature is valid or causal. Verify the cause with: Inspect Clarify attribution results, feature timestamp semantics, and whether the flag is created after the target event. Then confirm: The model no longer relies on future-only flags and held-out results reflect features available at inference. Reported signal: A tabular model shows one post-outcome administrative flag as its strongest feature in SageMaker Clarify despite implausibly high offline scores. Operational setting: At the model-governance gate before promotion of a new ML artifact.

Incorrect Answers

A:

Compare RMSE with MAE or robust error summaries and choose the metric matching the application's penalty for large misses. This addresses a different issue involving rmse versus mae outlier sensitivity. It does not correct the stated clarify feature attribution against target leakage mechanism. The targeted response is: Remove label-derived or post-outcome features, rebuild a point-in-time split, and recompute explanations and holdout metrics. Case evidence: A tabular model shows one post-outcome administrative flag as its strongest feature in SageMaker Clarify despite implausibly high offline scores. Operational setting: At the model-governance gate before promotion of a new ML artifact.

B:

Configure SageMaker Automatic Model Tuning with a bounded search space, objective metric, budget, and justified Bayesian or random strategy. This addresses a different issue involving hyperparameter bayesian tuning strategy. It does not correct the stated clarify feature attribution against target leakage mechanism. The targeted response is: Remove label-derived or post-outcome features, rebuild a point-in-time split, and recompute explanations and holdout metrics. Case evidence: A tabular model shows one post-outcome administrative flag as its strongest feature in SageMaker Clarify despite implausibly high offline scores. Operational setting: At the model-governance gate before promotion of a new ML artifact.

C:

Use a time-series forecasting approach with past context and seasonal features, evaluating all intended future horizons. This addresses a different issue involving time-series forecast horizon and seasonality. It does not correct the stated clarify feature attribution against target leakage mechanism. The targeted response is: Remove label-derived or post-outcome features, rebuild a point-in-time split, and recompute explanations and holdout metrics. Case evidence: A tabular model shows one post-outcome administrative flag as its strongest feature in SageMaker Clarify despite implausibly high offline scores. Operational setting: At the model-governance gate before promotion of a new ML artifact.

Question 84

During an operational load and release rehearsal for SageMaker-hosted inference, the following condition is observed: A real-time ML service is configured for very low baseline capacity but traffic can surge tenfold within seconds, exhausting available instance capacity. Change control requires measurable endpoint or pipeline acceptance evidence. Which measured outcome would confirm that the correction worked as intended?

  1. A synthetic load surge meets error and latency SLOs without rejected provisioning due to quota or startup lag.
  2. The build uploads signed artifacts successfully while unauthorized buckets and prefixes remain denied.
  3. The deployment loads artifacts privately and blocked public routes remain unused.
  4. Only the intended completed data batch launches one controlled training workflow, while unrelated objects do not.

Explanation

Correct Answer

A:

A synthetic load surge meets error and latency SLOs without rejected provisioning due to quota or startup lag. This is the acceptance evidence for the described endpoint scale-to-zero cold capacity planning scenario. Reactive scaling takes time and is bounded by quotas; zero spare capacity may not absorb an instantaneous burst. The corrective implementation it validates is: Set adequate minimum capacity, scaling limits, alarms, and request backpressure for the expected traffic ramp. Reported signal: A real-time ML service is configured for very low baseline capacity but traffic can surge tenfold within seconds, exhausting available instance capacity. Operational setting: During an operational load and release rehearsal for SageMaker-hosted inference.

Incorrect Answers

B:

The build uploads signed artifacts successfully while unauthorized buckets and prefixes remain denied. This could validate codebuild iam least-privilege artifact upload, but it does not establish that endpoint scale-to-zero cold capacity planning is resolved for this workload. Confirm instead: A synthetic load surge meets error and latency SLOs without rejected provisioning due to quota or startup lag. Observed problem: A real-time ML service is configured for very low baseline capacity but traffic can surge tenfold within seconds, exhausting available instance capacity. Operational setting: During an operational load and release rehearsal for SageMaker-hosted inference.

C:

The deployment loads artifacts privately and blocked public routes remain unused. This could validate sagemaker vpc endpoint access for private artifacts, but it does not establish that endpoint scale-to-zero cold capacity planning is resolved for this workload. Confirm instead: A synthetic load surge meets error and latency SLOs without rejected provisioning due to quota or startup lag. Observed problem: A real-time ML service is configured for very low baseline capacity but traffic can surge tenfold within seconds, exhausting available instance capacity. Operational setting: During an operational load and release rehearsal for SageMaker-hosted inference.

D:

Only the intended completed data batch launches one controlled training workflow, while unrelated objects do not. This could validate eventbridge conditional retraining trigger, but it does not establish that endpoint scale-to-zero cold capacity planning is resolved for this workload. Confirm instead: A synthetic load surge meets error and latency SLOs without rejected provisioning due to quota or startup lag. Observed problem: A real-time ML service is configured for very low baseline capacity but traffic can surge tenfold within seconds, exhausting available instance capacity. Operational setting: During an operational load and release rehearsal for SageMaker-hosted inference.

Question 85

At the model-governance gate before promotion of a new ML artifact, the following condition is observed: A deep model begins producing NaN loss midway through training and the team increases epochs instead of checking gradient stability. The team needs a defensible technical decision rather than an untracked workaround. What targeted change should be implemented after verifying the stated cause?

  1. Configure a supported SageMaker checkpoint location and resume training state on managed Spot interruption recovery.
  2. Remove label-derived or post-outcome features, rebuild a point-in-time split, and recompute explanations and holdout metrics.
  3. Select a licensed, task-appropriate model and validate its performance and resource needs before a controlled deployment.
  4. Use appropriate SageMaker Debugger or framework instrumentation; adjust learning rate or gradient clipping after confirming instability.

Explanation

Correct Answer

D:

Use appropriate SageMaker Debugger or framework instrumentation; adjust learning rate or gradient clipping after confirming instability. This is the appropriate controlled remedy for model debugger exploding gradient diagnosis. Divergent gradients can cause nonfinite optimization; more epochs do not repair an unstable update rule. Verify the cause with: Inspect tensor summaries, gradient magnitudes, learning-rate schedule, and training loss anomalies with supported debugging hooks. Then confirm: Gradient norms stay within the chosen range and training loss remains finite under representative batches. Reported signal: A deep model begins producing NaN loss midway through training and the team increases epochs instead of checking gradient stability. Operational setting: At the model-governance gate before promotion of a new ML artifact.

Incorrect Answers

A:

Configure a supported SageMaker checkpoint location and resume training state on managed Spot interruption recovery. This addresses a different issue involving managed spot training checkpoint recovery. It does not correct the stated model debugger exploding gradient diagnosis mechanism. The targeted response is: Use appropriate SageMaker Debugger or framework instrumentation; adjust learning rate or gradient clipping after confirming instability. Case evidence: A deep model begins producing NaN loss midway through training and the team increases epochs instead of checking gradient stability. Operational setting: At the model-governance gate before promotion of a new ML artifact.

B:

Remove label-derived or post-outcome features, rebuild a point-in-time split, and recompute explanations and holdout metrics. This addresses a different issue involving clarify feature attribution against target leakage. It does not correct the stated model debugger exploding gradient diagnosis mechanism. The targeted response is: Use appropriate SageMaker Debugger or framework instrumentation; adjust learning rate or gradient clipping after confirming instability. Case evidence: A deep model begins producing NaN loss midway through training and the team increases epochs instead of checking gradient stability. Operational setting: At the model-governance gate before promotion of a new ML artifact.

C:

Select a licensed, task-appropriate model and validate its performance and resource needs before a controlled deployment. This addresses a different issue involving sagemaker jumpstart template provenance. It does not correct the stated model debugger exploding gradient diagnosis mechanism. The targeted response is: Use appropriate SageMaker Debugger or framework instrumentation; adjust learning rate or gradient clipping after confirming instability. Case evidence: A deep model begins producing NaN loss midway through training and the team increases epochs instead of checking gradient stability. Operational setting: At the model-governance gate before promotion of a new ML artifact.

Question 86

At a production AWS ML data lake supporting customer-facing predictions, the following condition is observed: An S3 event-triggered feature ingestion function processes the same uploaded object twice, creating duplicate labeled samples. Engineers must avoid changing unrelated source partitions. Which measured outcome would confirm that the correction worked as intended?

  1. The report quantifies class or group imbalance with reproducible group counts and documents remediation assumptions.
  2. The transformed distribution is less skewed and inference uses identical guarded transformations.
  3. Replaying the same notification leaves exactly one corresponding logical training record and records deduplication.
  4. Both supported source versions produce the agreed typed feature, and invalid values enter an auditable reject path.

Explanation

Correct Answer

C:

Replaying the same notification leaves exactly one corresponding logical training record and records deduplication. This is the acceptance evidence for the described s3 object event at-least-once processing scenario. Event notifications can be delivered more than once; consuming an event is not proof of exactly-once effects. The corrective implementation it validates is: Make the ingestion handler idempotent using a stable object/version key and conditional ledger writes. Reported signal: An S3 event-triggered feature ingestion function processes the same uploaded object twice, creating duplicate labeled samples. Operational setting: At a production AWS ML data lake supporting customer-facing predictions.

Incorrect Answers

A:

The report quantifies class or group imbalance with reproducible group counts and documents remediation assumptions. This could validate clarify label imbalance measurement, but it does not establish that s3 object event at-least-once processing is resolved for this workload. Confirm instead: Replaying the same notification leaves exactly one corresponding logical training record and records deduplication. Observed problem: An S3 event-triggered feature ingestion function processes the same uploaded object twice, creating duplicate labeled samples. Operational setting: At a production AWS ML data lake supporting customer-facing predictions.

B:

The transformed distribution is less skewed and inference uses identical guarded transformations. This could validate skewed numerical feature log transformation, but it does not establish that s3 object event at-least-once processing is resolved for this workload. Confirm instead: Replaying the same notification leaves exactly one corresponding logical training record and records deduplication. Observed problem: An S3 event-triggered feature ingestion function processes the same uploaded object twice, creating duplicate labeled samples. Operational setting: At a production AWS ML data lake supporting customer-facing predictions.

D:

Both supported source versions produce the agreed typed feature, and invalid values enter an auditable reject path. This could validate glue schema evolution with mixed input types, but it does not establish that s3 object event at-least-once processing is resolved for this workload. Confirm instead: Replaying the same notification leaves exactly one corresponding logical training record and records deduplication. Observed problem: An S3 event-triggered feature ingestion function processes the same uploaded object twice, creating duplicate labeled samples. Operational setting: At a production AWS ML data lake supporting customer-facing predictions.

Question 87

As a regional data-pipeline rollout is validated against approved AWS storage and streams, the following condition is observed: A demand forecast model is evaluated on randomly sampled days while features contain recent historical moving averages, creating unrealistic future information. The data team must preserve event-time meaning and lineage. Which corrective action most directly addresses this ML workflow problem?

  1. Define explicit compatible conversions and schema contracts, isolate incompatible versions, and validate before publishing features.
  2. Place shared POSIX-style working files on suitable EFS or FSx storage and keep immutable large artifacts in S3.
  3. Measure label distribution differences with an appropriate pretraining bias metric and improve data representation before training.
  4. Use rolling-origin or chronological validation with strictly past-derived features and appropriate forecast horizons.

Explanation

Correct Answer

D:

Use rolling-origin or chronological validation with strictly past-derived features and appropriate forecast horizons. This is the appropriate controlled remedy for chronological split for forecast evaluation. Random splits violate temporal causality when future observations influence features or train data. Verify the cause with: Inspect feature availability timestamps and whether each label time follows the last feature observation. Then confirm: Every held-out forecast uses only data that existed at prediction time and its backtest follows real ordering. Reported signal: A demand forecast model is evaluated on randomly sampled days while features contain recent historical moving averages, creating unrealistic future information. Operational setting: As a regional data-pipeline rollout is validated against approved AWS storage and streams.

Incorrect Answers

A:

Define explicit compatible conversions and schema contracts, isolate incompatible versions, and validate before publishing features. This addresses a different issue involving glue schema evolution with mixed input types. It does not correct the stated chronological split for forecast evaluation mechanism. The targeted response is: Use rolling-origin or chronological validation with strictly past-derived features and appropriate forecast horizons. Case evidence: A demand forecast model is evaluated on randomly sampled days while features contain recent historical moving averages, creating unrealistic future information. Operational setting: As a regional data-pipeline rollout is validated against approved AWS storage and streams.

B:

Place shared POSIX-style working files on suitable EFS or FSx storage and keep immutable large artifacts in S3. This addresses a different issue involving s3 versus efs training storage choice. It does not correct the stated chronological split for forecast evaluation mechanism. The targeted response is: Use rolling-origin or chronological validation with strictly past-derived features and appropriate forecast horizons. Case evidence: A demand forecast model is evaluated on randomly sampled days while features contain recent historical moving averages, creating unrealistic future information. Operational setting: As a regional data-pipeline rollout is validated against approved AWS storage and streams.

C:

Measure label distribution differences with an appropriate pretraining bias metric and improve data representation before training. This addresses a different issue involving clarify label imbalance measurement. It does not correct the stated chronological split for forecast evaluation mechanism. The targeted response is: Use rolling-origin or chronological validation with strictly past-derived features and appropriate forecast horizons. Case evidence: A demand forecast model is evaluated on randomly sampled days while features contain recent historical moving averages, creating unrealistic future information. Operational setting: As a regional data-pipeline rollout is validated against approved AWS storage and streams.

Question 88

At a production AWS ML data lake supporting customer-facing predictions, the following condition is observed: A Spark transformation joins billions of records on a region key, and one region contains most observations, stalling one shuffle task. Engineers must avoid changing unrelated source partitions. Which engineering response resolves the actual issue without masking the symptoms?

  1. Drop leakage-prone IDs or adopt controlled hashing or suitable learned encodings with stable unknown-category handling.
  2. Repartition or salt the skewed join appropriately, or use a justified broadcast join when the small side fits memory.
  3. Apply resampling only inside each training fold and evaluate untouched validation folds using precision and recall metrics.
  4. Choose partition keys with enough cardinality or reshard the stream for the intended throughput, while preserving required per-key ordering.

Explanation

Correct Answer

B:

Repartition or salt the skewed join appropriately, or use a justified broadcast join when the small side fits memory. This is the appropriate controlled remedy for glue spark join skew on dominant identifier. A low-cardinality or skewed key can concentrate a distributed join into one expensive partition. Verify the cause with: Inspect Spark stage durations, partition size distribution, join key frequencies, and executor spill metrics. Then confirm: Long-tail task duration and shuffle spill decline without changing the joined record counts. Reported signal: A Spark transformation joins billions of records on a region key, and one region contains most observations, stalling one shuffle task. Operational setting: At a production AWS ML data lake supporting customer-facing predictions.

Incorrect Answers

A:

Drop leakage-prone IDs or adopt controlled hashing or suitable learned encodings with stable unknown-category handling. This addresses a different issue involving high-cardinality categorical encoding. It does not correct the stated glue spark join skew on dominant identifier mechanism. The targeted response is: Repartition or salt the skewed join appropriately, or use a justified broadcast join when the small side fits memory. Case evidence: A Spark transformation joins billions of records on a region key, and one region contains most observations, stalling one shuffle task. Operational setting: At a production AWS ML data lake supporting customer-facing predictions.

C:

Apply resampling only inside each training fold and evaluate untouched validation folds using precision and recall metrics. This addresses a different issue involving class imbalance handled inside cross-validation. It does not correct the stated glue spark join skew on dominant identifier mechanism. The targeted response is: Repartition or salt the skewed join appropriately, or use a justified broadcast join when the small side fits memory. Case evidence: A Spark transformation joins billions of records on a region key, and one region contains most observations, stalling one shuffle task. Operational setting: At a production AWS ML data lake supporting customer-facing predictions.

D:

Choose partition keys with enough cardinality or reshard the stream for the intended throughput, while preserving required per-key ordering. This addresses a different issue involving kinesis data streams shard hot key. It does not correct the stated glue spark join skew on dominant identifier mechanism. The targeted response is: Repartition or salt the skewed join appropriately, or use a justified broadcast join when the small side fits memory. Case evidence: A Spark transformation joins billions of records on a region key, and one region contains most observations, stalling one shuffle task. Operational setting: At a production AWS ML data lake supporting customer-facing predictions.

Question 89

As a regional data-pipeline rollout is validated against approved AWS storage and streams, the following condition is observed: Several distributed workers repeatedly read small shared checkpoints with POSIX file semantics, but the pipeline treats object-storage paths as a mutable shared filesystem. The data team must preserve event-time meaning and lineage. Which corrective action most directly addresses this ML workflow problem?

  1. Use rolling-origin or chronological validation with strictly past-derived features and appropriate forecast horizons.
  2. Increase appropriate retention before long maintenance and persist a durable downstream historical copy for replay beyond retention.
  3. Place shared POSIX-style working files on suitable EFS or FSx storage and keep immutable large artifacts in S3.
  4. Drop leakage-prone IDs or adopt controlled hashing or suitable learned encodings with stable unknown-category handling.

Explanation

Correct Answer

C:

Place shared POSIX-style working files on suitable EFS or FSx storage and keep immutable large artifacts in S3. This is the appropriate controlled remedy for s3 versus efs training storage choice. S3 object semantics differ from mounted shared filesystems; choose storage for workload access behavior. Verify the cause with: Inspect workload read/write semantics, shared access requirements, filesystem operations, and object-store request patterns. Then confirm: Concurrent workers observe consistent supported shared-file operations and training artifact delivery remains durable. Reported signal: Several distributed workers repeatedly read small shared checkpoints with POSIX file semantics, but the pipeline treats object-storage paths as a mutable shared filesystem. Operational setting: As a regional data-pipeline rollout is validated against approved AWS storage and streams.

Incorrect Answers

A:

Use rolling-origin or chronological validation with strictly past-derived features and appropriate forecast horizons. This addresses a different issue involving chronological split for forecast evaluation. It does not correct the stated s3 versus efs training storage choice mechanism. The targeted response is: Place shared POSIX-style working files on suitable EFS or FSx storage and keep immutable large artifacts in S3. Case evidence: Several distributed workers repeatedly read small shared checkpoints with POSIX file semantics, but the pipeline treats object-storage paths as a mutable shared filesystem. Operational setting: As a regional data-pipeline rollout is validated against approved AWS storage and streams.

B:

Increase appropriate retention before long maintenance and persist a durable downstream historical copy for replay beyond retention. This addresses a different issue involving kinesis consumer retention gap. It does not correct the stated s3 versus efs training storage choice mechanism. The targeted response is: Place shared POSIX-style working files on suitable EFS or FSx storage and keep immutable large artifacts in S3. Case evidence: Several distributed workers repeatedly read small shared checkpoints with POSIX file semantics, but the pipeline treats object-storage paths as a mutable shared filesystem. Operational setting: As a regional data-pipeline rollout is validated against approved AWS storage and streams.

D:

Drop leakage-prone IDs or adopt controlled hashing or suitable learned encodings with stable unknown-category handling. This addresses a different issue involving high-cardinality categorical encoding. It does not correct the stated s3 versus efs training storage choice mechanism. The targeted response is: Place shared POSIX-style working files on suitable EFS or FSx storage and keep immutable large artifacts in S3. Case evidence: Several distributed workers repeatedly read small shared checkpoints with POSIX file semantics, but the pipeline treats object-storage paths as a mutable shared filesystem. Operational setting: As a regional data-pipeline rollout is validated against approved AWS storage and streams.

Question 90

While reproducing the alert in an isolated AWS ML integration account, the following condition is observed: An ML pipeline in one account is allowed to write model predictions to an unapproved external bucket because its role has overly broad s3:PutObject access. The proposed change must preserve logging and least privilege. Which measured outcome would confirm that the correction worked as intended?

  1. A test write to an approved prefix succeeds, while writes to unapproved external destinations are denied.
  2. Changed feature distributions produce attributable violations even while endpoint availability remains healthy.
  3. The endpoint loads the intended encrypted artifact while unrelated principals cannot decrypt it.
  4. Repeated requests for one test user reach the same assigned variant and measured outcomes remain comparable.

Explanation

Correct Answer

A:

A test write to an approved prefix succeeds, while writes to unapproved external destinations are denied. This is the acceptance evidence for the described cross-account s3 bucket exfiltration protection scenario. Least privilege must constrain the destination resource; an unrestricted object-write policy can enable data exfiltration. The corrective implementation it validates is: Restrict writes to approved bucket prefixes and apply appropriate resource and organizational guardrails. Reported signal: An ML pipeline in one account is allowed to write model predictions to an unapproved external bucket because its role has overly broad s3:PutObject access. Operational setting: While reproducing the alert in an isolated AWS ML integration account.

Incorrect Answers

B:

Changed feature distributions produce attributable violations even while endpoint availability remains healthy. This could validate sagemaker model monitor feature drift baseline, but it does not establish that cross-account s3 bucket exfiltration protection is resolved for this workload. Confirm instead: A test write to an approved prefix succeeds, while writes to unapproved external destinations are denied. Observed problem: An ML pipeline in one account is allowed to write model predictions to an unapproved external bucket because its role has overly broad s3:PutObject access. Operational setting: While reproducing the alert in an isolated AWS ML integration account.

C:

The endpoint loads the intended encrypted artifact while unrelated principals cannot decrypt it. This could validate s3 sse-kms cross-account model artifact access, but it does not establish that cross-account s3 bucket exfiltration protection is resolved for this workload. Confirm instead: A test write to an approved prefix succeeds, while writes to unapproved external destinations are denied. Observed problem: An ML pipeline in one account is allowed to write model predictions to an unapproved external bucket because its role has overly broad s3:PutObject access. Operational setting: While reproducing the alert in an isolated AWS ML integration account.

D:

Repeated requests for one test user reach the same assigned variant and measured outcomes remain comparable. This could validate champion challenger a/b assignment consistency, but it does not establish that cross-account s3 bucket exfiltration protection is resolved for this workload. Confirm instead: A test write to an approved prefix succeeds, while writes to unapproved external destinations are denied. Observed problem: An ML pipeline in one account is allowed to write model predictions to an unapproved external bucket because its role has overly broad s3:PutObject access. Operational setting: While reproducing the alert in an isolated AWS ML integration account.

AWS Certified Machine Learning Engineer - Associate MLA-C01 Practice Tests