Exam Bundle

Deduplicate on stable business identity and event-version rules, retaining the latest accepted entity state as required. It does not directly address the Granite SaaS case: A scheduled training pipeline reads CSV from an S3 bucket, but individual files contain inconsistent delimiters and the parsing job silently shifts feature columns. The relevant acceptance criterion for this case is: A controlled malformed sample enters the rejected-record path while valid rows retain their expected column names and feature values.

Apply an appropriate categorical encoding while preserving the region identifier's nominal semantics and missing-value policy. It does not directly address the Granite SaaS case: A scheduled training pipeline reads CSV from an S3 bucket, but individual files contain inconsistent delimiters and the parsing job silently shifts feature columns. The relevant acceptance criterion for this case is: A controlled malformed sample enters the rejected-record path while valid rows retain their expected column names and feature values.

Use a stratified or otherwise class-aware splitting method with declared minimum representation checks for the holdout set. It does not directly address the Granite SaaS case: A scheduled training pipeline reads CSV from an S3 bucket, but individual files contain inconsistent delimiters and the parsing job silently shifts feature columns. The relevant acceptance criterion for this case is: A controlled malformed sample enters the rejected-record path while valid rows retain their expected column names and feature values.

A deliberately labeled test row never exposes its target through the predictor schema and the suspicious accuracy is re-evaluated. This would verify a different operation: Implement a schema-level separation of label and predictor columns and verify feature lineage prior to fitting the model. That separate action addresses: A data scientist encodes the target label as an input feature for convenience and obtains a nearly perfect training score. It cannot establish the required Granite SaaS correction for: A scheduled training pipeline reads CSV from an S3 bucket, but individual files contain inconsistent delimiters and the parsing job silently shifts feature columns. The decisive acceptance check is instead: A controlled malformed sample enters the rejected-record path while valid rows retain their expected column names and feature values.

A controlled rebuild with the same recipe and source manifest reproduces feature checksums and row counts. This would verify a different operation: Record and deploy a specific reviewed recipe version together with input snapshot identifiers for every training dataset build. That separate action addresses: An AWS Glue DataBrew recipe is edited directly in production and a rerun cannot reproduce the original training features. It cannot establish the required Granite SaaS correction for: A scheduled training pipeline reads CSV from an S3 bucket, but individual files contain inconsistent delimiters and the parsing job silently shifts feature columns. The decisive acceptance check is instead: A controlled malformed sample enters the rejected-record path while valid rows retain their expected column names and feature values.

Every required rare class appears at least the configured minimum count in evaluation and no training row is duplicated into test. This would verify a different operation: Use a stratified or otherwise class-aware splitting method with declared minimum representation checks for the holdout set. That separate action addresses: A multi-class label dataset is split without stratification and a rare class is missing entirely from the evaluation partition. It cannot establish the required Granite SaaS correction for: A scheduled training pipeline reads CSV from an S3 bucket, but individual files contain inconsistent delimiters and the parsing job silently shifts feature columns. The decisive acceptance check is instead: A controlled malformed sample enters the rejected-record path while valid rows retain their expected column names and feature values.

Fit and persist a categorical encoder on the training set with an explicit unknown-category handling policy for all later inputs. It does not directly address the Westhaven Retail case: A scheduled training pipeline reads CSV from an S3 bucket, but individual files contain inconsistent delimiters and the parsing job silently shifts feature columns. The relevant acceptance criterion for this case is: A controlled malformed sample enters the rejected-record path while valid rows retain their expected column names and feature values.

Attach source system, ingest time and immutable batch manifest metadata to accepted training records during data integration. It does not directly address the Westhaven Retail case: A scheduled training pipeline reads CSV from an S3 bucket, but individual files contain inconsistent delimiters and the parsing job silently shifts feature columns. The relevant acceptance criterion for this case is: A controlled malformed sample enters the rejected-record path while valid rows retain their expected column names and feature values.

Fit a standardization or scaling transform on training data only and apply the saved transform to validation and inference inputs. It does not directly address the Westhaven Retail case: A scheduled training pipeline reads CSV from an S3 bucket, but individual files contain inconsistent delimiters and the parsing job silently shifts feature columns. The relevant acceptance criterion for this case is: A controlled malformed sample enters the rejected-record path while valid rows retain their expected column names and feature values.

Zero and positive test values transform into finite documented values and the resulting feature remains suitable for the model. This would verify a different operation: Choose a guarded nonnegative log transformation and test boundary handling before saving the feature pipeline. That separate action addresses: An ML regression dataset has a heavily skewed positive feature, but an unguarded logarithmic transform generates invalid values for zero entries. It cannot establish the required Westhaven Retail correction for: A scheduled training pipeline reads CSV from an S3 bucket, but individual files contain inconsistent delimiters and the parsing job silently shifts feature columns. The decisive acceptance check is instead: A controlled malformed sample enters the rejected-record path while valid rows retain their expected column names and feature values.

A deliberately labeled test row never exposes its target through the predictor schema and the suspicious accuracy is re-evaluated. This would verify a different operation: Implement a schema-level separation of label and predictor columns and verify feature lineage prior to fitting the model. That separate action addresses: A data scientist encodes the target label as an input feature for convenience and obtains a nearly perfect training score. It cannot establish the required Westhaven Retail correction for: A scheduled training pipeline reads CSV from an S3 bucket, but individual files contain inconsistent delimiters and the parsing job silently shifts feature columns. The decisive acceptance check is instead: A controlled malformed sample enters the rejected-record path while valid rows retain their expected column names and feature values.

The large synthetic partition completes without driver out-of-memory errors and its aggregate counts match the reference run. This would verify a different operation: Use distributed Spark transforms or scalable partitioned outputs instead of collecting the full dataset to one process. That separate action addresses: A Spark feature preparation stage converts all encoded records to driver memory and fails for large training partitions. It cannot establish the required Westhaven Retail correction for: A scheduled training pipeline reads CSV from an S3 bucket, but individual files contain inconsistent delimiters and the parsing job silently shifts feature columns. The decisive acceptance check is instead: A controlled malformed sample enters the rejected-record path while valid rows retain their expected column names and feature values.

Validate nullable field semantics and reject or impute explicit missing tokens without converting them to legitimate zero values. It does not directly address the Brookside Manufacturing case: An event-driven fraud model receives new records in Kinesis, but the consumer uses a hard-coded shard checkpoint that replays the same events after restart. The relevant acceptance criterion for this case is: A forced consumer restart resumes from the committed position and produces no duplicated training events for the test partition.

Make the feature write idempotent using event identifiers or version checks and send repeated failures to a controlled recovery path. It does not directly address the Brookside Manufacturing case: An event-driven fraud model receives new records in Kinesis, but the consumer uses a hard-coded shard checkpoint that replays the same events after restart. The relevant acceptance criterion for this case is: A forced consumer restart resumes from the committed position and produces no duplicated training events for the test partition.

Split the original records first, then apply synthetic oversampling exclusively to the training split. It does not directly address the Brookside Manufacturing case: An event-driven fraud model receives new records in Kinesis, but the consumer uses a hard-coded shard checkpoint that replays the same events after restart. The relevant acceptance criterion for this case is: A forced consumer restart resumes from the committed position and produces no duplicated training events for the test partition.

A scheduled retention test can still retrieve every manifest object needed to reproduce the archived model version. This would verify a different operation: Apply retention and lifecycle rules compatible with artifact lineage and archive immutable source objects required for approved reproducibility. That separate action addresses: A training manifest refers to S3 objects that were removed by an aggressive lifecycle rule, so reproducible model builds fail months later. It cannot establish the required Brookside Manufacturing correction for: An event-driven fraud model receives new records in Kinesis, but the consumer uses a hard-coded shard checkpoint that replays the same events after restart. The decisive acceptance check is instead: A forced consumer restart resumes from the committed position and produces no duplicated training events for the test partition.

A multilingual test phrase yields the same token IDs in the training pipeline and deployed inference endpoint. This would verify a different operation: Version a single canonical text normalization and tokenizer artifact and use it in both processing and serving containers. That separate action addresses: A text classifier tokenizes raw Unicode text differently in batch training and the real-time inference application. It cannot establish the required Brookside Manufacturing correction for: An event-driven fraud model receives new records in Kinesis, but the consumer uses a hard-coded shard checkpoint that replays the same events after restart. The decisive acceptance check is instead: A forced consumer restart resumes from the committed position and produces no duplicated training events for the test partition.

A deliberately labeled test row never exposes its target through the predictor schema and the suspicious accuracy is re-evaluated. This would verify a different operation: Implement a schema-level separation of label and predictor columns and verify feature lineage prior to fitting the model. That separate action addresses: A data scientist encodes the target label as an input feature for convenience and obtains a nearly perfect training score. It cannot establish the required Brookside Manufacturing correction for: An event-driven fraud model receives new records in Kinesis, but the consumer uses a hard-coded shard checkpoint that replays the same events after restart. The decisive acceptance check is instead: A forced consumer restart resumes from the committed position and produces no duplicated training events for the test partition.

Learn imputation statistics exclusively from the training split and reuse those fixed values on held-out datasets. It does not directly address the Bluehaven Transit case: An event-driven fraud model receives new records in Kinesis, but the consumer uses a hard-coded shard checkpoint that replays the same events after restart. The relevant acceptance criterion for this case is: A forced consumer restart resumes from the committed position and produces no duplicated training events for the test partition.

Materialize governed versioned features with an appropriate offline store and documented recomputation schedule. It does not directly address the Bluehaven Transit case: An event-driven fraud model receives new records in Kinesis, but the consumer uses a hard-coded shard checkpoint that replays the same events after restart. The relevant acceptance criterion for this case is: A forced consumer restart resumes from the committed position and produces no duplicated training events for the test partition.

Version a single canonical text normalization and tokenizer artifact and use it in both processing and serving containers. It does not directly address the Bluehaven Transit case: An event-driven fraud model receives new records in Kinesis, but the consumer uses a hard-coded shard checkpoint that replays the same events after restart. The relevant acceptance criterion for this case is: A forced consumer restart resumes from the committed position and produces no duplicated training events for the test partition.

A scheduled retention test can still retrieve every manifest object needed to reproduce the archived model version. This would verify a different operation: Apply retention and lifecycle rules compatible with artifact lineage and archive immutable source objects required for approved reproducibility. That separate action addresses: A training manifest refers to S3 objects that were removed by an aggressive lifecycle rule, so reproducible model builds fail months later. It cannot establish the required Bluehaven Transit correction for: An event-driven fraud model receives new records in Kinesis, but the consumer uses a hard-coded shard checkpoint that replays the same events after restart. The decisive acceptance check is instead: A forced consumer restart resumes from the committed position and produces no duplicated training events for the test partition.

A bilingual reference set retains the required product terms and its label agreement remains inside the approved quality range. This would verify a different operation: Preserve protected product tokens or use a validated multilingual feature representation that does not corrupt business identifiers. That separate action addresses: A sentiment pipeline blindly translates all customer messages into English even though translation changes domain-specific product identifiers. It cannot establish the required Bluehaven Transit correction for: An event-driven fraud model receives new records in Kinesis, but the consumer uses a hard-coded shard checkpoint that replays the same events after restart. The decisive acceptance check is instead: A forced consumer restart resumes from the committed position and produces no duplicated training events for the test partition.

Distance-sensitive evaluation improves and the identical stored scale parameters are used during training and online inference. This would verify a different operation: Fit a standardization or scaling transform on training data only and apply the saved transform to validation and inference inputs. That separate action addresses: A churn model trains on customer age and transaction amount with unrelated scales, causing its distance-based feature representation to be dominated by transaction magnitude. It cannot establish the required Bluehaven Transit correction for: An event-driven fraud model receives new records in Kinesis, but the consumer uses a hard-coded shard checkpoint that replays the same events after restart. The decisive acceptance check is instead: A forced consumer restart resumes from the committed position and produces no duplicated training events for the test partition.

Classify regional attributes and keep processing and storage in approved AWS Regions with authorized data residency controls. It does not directly address the Northgate Telecom case: A data engineering job stores a large column-oriented training dataset as thousands of verbose JSON files and suffers excessive scan costs in Athena. The relevant acceptance criterion for this case is: The benchmark query scans fewer bytes and still returns identical approved training rows and feature values.

Deduplicate on stable business identity and event-version rules, retaining the latest accepted entity state as required. It does not directly address the Northgate Telecom case: A data engineering job stores a large column-oriented training dataset as thousands of verbose JSON files and suffers excessive scan costs in Athena. The relevant acceptance criterion for this case is: The benchmark query scans fewer bytes and still returns identical approved training rows and feature values.

Use a suitable embedding or hashing approach with collision and memory impact assessed for the model objective. It does not directly address the Northgate Telecom case: A data engineering job stores a large column-oriented training dataset as thousands of verbose JSON files and suffers excessive scan costs in Athena. The relevant acceptance criterion for this case is: The benchmark query scans fewer bytes and still returns identical approved training rows and feature values.

A controlled rebuild with the same recipe and source manifest reproduces feature checksums and row counts. This would verify a different operation: Record and deploy a specific reviewed recipe version together with input snapshot identifiers for every training dataset build. That separate action addresses: An AWS Glue DataBrew recipe is edited directly in production and a rerun cannot reproduce the original training features. It cannot establish the required Northgate Telecom correction for: A data engineering job stores a large column-oriented training dataset as thousands of verbose JSON files and suffers excessive scan costs in Athena. The decisive acceptance check is instead: The benchmark query scans fewer bytes and still returns identical approved training rows and feature values.

The final report records positive-class recall and precision on the original distribution, not only inflated aggregate accuracy. This would verify a different operation: Measure class balance and compare stratified resampling or class weighting on training data while preserving an untouched evaluation set. That separate action addresses: A marketing propensity model has a class distribution of 98 percent negative examples and selection by accuracy alone hides poor positive detection. It cannot establish the required Northgate Telecom correction for: A data engineering job stores a large column-oriented training dataset as thousands of verbose JSON files and suffers excessive scan costs in Athena. The decisive acceptance check is instead: The benchmark query scans fewer bytes and still returns identical approved training rows and feature values.

An input set with mixed-capitalization labels maps consistently while an unknown label triggers the declared quality exception. This would verify a different operation: Define a validated label encoding dictionary and reject unmapped target values before model training starts. That separate action addresses: A binary target is stored as strings with inconsistent capitalization and preprocessing silently maps one valid positive class to negative. It cannot establish the required Northgate Telecom correction for: A data engineering job stores a large column-oriented training dataset as thousands of verbose JSON files and suffers excessive scan costs in Athena. The decisive acceptance check is instead: The benchmark query scans fewer bytes and still returns identical approved training rows and feature values.