Databricks Certified Machine Learning Professional Practice Test 88
Question 870
The engineering team encounters the following issue: An engineer stores model weights in a personal workspace folder and treats that path as a production artifact registry with audit history. What direct consequence should the owner prioritize if no fix is made?
- Production inference can access unrelated protected data or train on unintended tables.
- The artifact can disappear with an account change or be overwritten without reliable provenance.
- A healthy deployed job crashes on missing fields or silently reads wrong defaults.
- The live model logic diverges from tested code and cannot be reproduced during an incident.
Explanation
Correct Answer
B:This consequence follows from the specific unresolved condition. The artifact can disappear with an account change or be overwritten without reliable provenance. The causal constraint is: Governed model registration and access controls provide traceability beyond an ad hoc personal file path.
Incorrect Answers
A:That direct consequence is directed at a different model lifecycle management situation: A team trains on curated development features but deploys the same job with credentials that accidentally read unapproved production source tables. The more immediate risk here is the artifact can disappear with an account change or be overwritten without reliable provenance.
C:That direct consequence is directed at a different model lifecycle management situation: A customer segmentation project promotes a scoring job but omits a database migration required by its new feature schema. The more immediate risk here is the artifact can disappear with an account change or be overwritten without reliable provenance.
D:That direct consequence is directed at a different model lifecycle management situation: A regulated company changes deployed Python notebook code directly in production to correct a model transform while its test environment runs a different untracked revision. The more immediate risk here is the artifact can disappear with an account change or be overwritten without reliable provenance.
Question 871
In an enterprise ML workflow, the reported issue is: A new recommendation model consumes three times as much memory and the team plans a canary based solely on prediction accuracy, ignoring serving capacity. How should the engineer validate the correction?
- Sampled live requests and endpoint configuration confirm candidate traffic is removed or bounded as approved.
- The candidate remains within memory and latency headroom at each authorized traffic increment.
- A readiness probe and known inference fixture both succeed before planned traffic movement.
- A release rehearsal restores champion predictions using the saved model/version and configuration.
Explanation
Correct Answer
B:This verification measures the required result rather than just process liveness. The candidate remains within memory and latency headroom at each authorized traffic increment. It confirms whether the targeted action succeeded: require resource and response-time acceptance criteria before expanding canary traffic.
Incorrect Answers
A:That acceptance test is directed at a different deployment strategies situation: During a controlled release, the candidate model's P95 latency triples and engineers want to change the registered model alias without adjusting endpoint routing. Here the actual success criterion is the candidate remains within memory and latency headroom at each authorized traffic increment.
C:That acceptance test is directed at a different deployment strategies situation: An endpoint update request succeeds but the newly configured model version is still loading; monitoring marks the release complete immediately. Here the actual success criterion is the candidate remains within memory and latency headroom at each authorized traffic increment.
D:That acceptance test is directed at a different deployment strategies situation: An ML operator promotes a new model version but its automated rollback points to a model that was deleted from the endpoint configuration. Here the actual success criterion is the candidate remains within memory and latency headroom at each authorized traffic increment.
Question 872
Review this production model incident: An Optuna optimization reuses the same MLflow run ID for concurrent trials, causing overlapping parameter values and artifacts. Which failure risk follows most closely from the unresolved behavior?
- Multiple trials contend for the same accelerator and fail with memory exhaustion.
- Training stalls on memory pressure even while unrelated cluster nodes are underutilized.
- One hot customer group repeatedly crashes tasks and stalls distributed tuning.
- A trial's metrics and model artifacts overwrite evidence from another configuration.
Explanation
Correct Answer
D:This consequence follows from the specific unresolved condition. A trial's metrics and model artifacts overwrite evidence from another configuration. The causal constraint is: Independent trial logs require distinct run identity or an explicit supported nested-run association under concurrency.
Incorrect Answers
A:That direct consequence is directed at a different scaling and tuning situation: A Ray Tune training function requests zero GPU resources but initializes a large GPU model inside every trial. The more immediate risk here is a trial's metrics and model artifacts overwrite evidence from another configuration.
B:That direct consequence is directed at a different scaling and tuning situation: A single-node model training run spends almost all time swapping a 40 GB feature matrix on a 16 GB VM; no distributed version exists. The more immediate risk here is a trial's metrics and model artifacts overwrite evidence from another configuration.
C:That direct consequence is directed at a different scaling and tuning situation: A recommendation team wraps a large pandas UDF around entire customer histories, then discovers some partitions exceed worker memory while others finish quickly. The more immediate risk here is a trial's metrics and model artifacts overwrite evidence from another configuration.
Question 873
Consider the following fictional Databricks production case: A fraud training row is labeled on March 10 but an ordinary entity join attaches an account balance updated on March 12. Which observation most directly tests the suspected failure boundary?
- Compare event timestamps, ingestion timestamps and window assignment for late synthetic readings.
- Inspect feature update timestamps against label timestamps for a sampled joined row.
- Measure endpoint latency attributable to offline table reads versus approved online lookup latency.
- Map feature consumers, source commit IDs and before/after results for a fixed regression fixture.
Explanation
Correct Answer
B:This is the most discriminating check for the observed symptoms. Inspect feature update timestamps against label timestamps for a sampled joined row. The underlying technical issue is that point-in-time training joins must select features available at or before each label observation, not the latest record irrespective of date.
Incorrect Answers
A:That diagnostic evidence is directed at a different advanced feature store concepts situation: A sensor feature pipeline computes rolling 30-minute averages by arrival time, although delayed events carry trustworthy event timestamps. For the stated incident, the decisive observation is inspect feature update timestamps against label timestamps for a sampled joined row.
C:That diagnostic evidence is directed at a different advanced feature store concepts situation: A pricing endpoint queries a large offline Delta feature table for every user request, and P95 response time reaches several seconds. For the stated incident, the decisive observation is inspect feature update timestamps against label timestamps for a sampled joined row.
D:That diagnostic evidence is directed at a different advanced feature store concepts situation: A team updates a shared on-demand feature function without version control and unexpectedly changes both batch retraining and online serving behavior. For the stated incident, the decisive observation is inspect feature update timestamps against label timestamps for a sampled joined row.
Question 874
In an enterprise ML workflow, the reported issue is: A team has accurate inference logs but its monitoring job never refreshes computed drift metrics after a table schema change. How should the engineer validate the correction?
- Synthetic daily shifts appear in separate time-series windows with correct timestamp boundaries.
- A newly ingested fixture advances metric refresh and produces an appropriately timestamped drift result.
- A safe synthetic threshold crossing reaches the correct team with the monitored metric and window.
- A scheduled profile run writes metrics under the approved identity without unauthorized table access.
Explanation
Correct Answer
B:This verification measures the required result rather than just process liveness. A newly ingested fixture advances metric refresh and produces an appropriately timestamped drift result. It confirms whether the targeted action succeeded: repair the profile computation and alert on stale metric output separately from the model's predictions.
Incorrect Answers
A:That acceptance test is directed at a different drift detection and lakehouse monitoring situation: A fleet sensor table records measurement timestamps and analysts need daily drift trends; a snapshot-only profile erases the time dimension. Here the actual success criterion is a newly ingested fixture advances metric refresh and produces an appropriately timestamped drift result.
C:That acceptance test is directed at a different drift detection and lakehouse monitoring situation: A fraud-feature drift monitor crosses its threshold but the alert has no assigned recipient and no incident ownership. Here the actual success criterion is a newly ingested fixture advances metric refresh and produces an appropriately timestamped drift result.
D:That acceptance test is directed at a different drift detection and lakehouse monitoring situation: A monitor is pointed at a Unity Catalog table but the automation identity lacks the privileges needed to generate profile outputs. Here the actual success criterion is a newly ingested fixture advances metric refresh and produces an appropriately timestamped drift result.
Question 875
A production ML review contains this situation: An analyst loads a fitted LogisticRegressionModel but calls fit on it for every scored batch rather than transform. Which check shows that this particular problem has been resolved?
- A held-out review confirms alert volume and true-positive capture at the frozen threshold.
- The full fitted pipeline completes on a small test DataFrame and exposes the planned input and output columns.
- The same versioned fitted model produces reproducible predictions for the saved scoring fixture.
- The training trace identifies worker-side execution and supported framework behavior for the selected model.
Explanation
Correct Answer
C:This verification measures the required result rather than just process liveness. The same versioned fitted model produces reproducible predictions for the saved scoring fixture. It confirms whether the targeted action succeeded: use model.transform on new feature vectors and retrain only in an approved training workflow.
Incorrect Answers
A:That acceptance test is directed at a different using spark ml situation: A binary classifier ranks risky cases well but sends too many alerts at a default 0.5 decision threshold. Here the actual success criterion is the same versioned fitted model produces reproducible predictions for the saved scoring fixture.
B:That acceptance test is directed at a different using spark ml situation: A pipeline engineer fits a vectorizer after an estimator in the stage list and then tries to use the estimator's prediction as the vectorizer input. Here the actual success criterion is the same versioned fitted model produces reproducible predictions for the saved scoring fixture.
D:That acceptance test is directed at a different using spark ml situation: A media group has a specialized deep neural model that is not implemented as a SparkML estimator and assumes SparkML will automatically distribute its GPU training. Here the actual success criterion is the same versioned fitted model produces reproducible predictions for the saved scoring fixture.
Question 876
The engineering team encounters the following issue: A model's prediction error appears to spike at 2 a.m. local time but the monitoring table groups timestamps in UTC for a globally distributed region. What corrective action should the engineer prefer?
- Normalize timestamp semantics before applying approved local-hour performance slices.
- Monitor leading demand indicators and configure supported capacity policies for predictable spikes.
- Calculate slice-specific error rates and a documented derived disparity metric with privacy safeguards.
- Compute comparable version-specific performance trends and retain traffic-share context.
Explanation
Correct Answer
A:This corrects the relevant failure boundary instead of merely hiding the symptom. Normalize timestamp semantics before applying approved local-hour performance slices. The issue matters because an apparent overnight model failure is only a grouping artifact and triggers costly false alerts.
Incorrect Answers
B:That corrective action is directed at a different drift detection and lakehouse monitoring situation: An endpoint receives a large request burst every morning and its monitoring alert reacts only to 5xx errors after autoscaling cannot keep up. The relevant corrective action here is normalize timestamp semantics before applying approved local-hour performance slices.
C:That corrective action is directed at a different drift detection and lakehouse monitoring situation: A fairness reviewer needs the difference in false-negative rates between two approved cohorts, but the monitor computes an aggregate without cohort labels. The relevant corrective action here is normalize timestamp semantics before applying approved local-hour performance slices.
D:That corrective action is directed at a different drift detection and lakehouse monitoring situation: A scoring endpoint quietly moves from model v7 to v8 during a rollout but an inference trend aggregates both versions in one error curve. The relevant corrective action here is normalize timestamp semantics before applying approved local-hour performance slices.
Question 877
The engineering team encounters the following issue: An engineer uses only production replay data to validate a code change and exposes customer events in an unrestricted test workspace. Which experiment most directly demonstrates a successful fix?
- All six consumers use one pinned helper version and identical threshold fixtures.
- The same event time is interpreted consistently at every persisted lifecycle boundary.
- The unit suite confirms known instants normalize consistently across supported offset formats.
- Test runs require no unauthorized identifiers and negative access probes fail as intended.
Explanation
Correct Answer
D:This verification measures the required result rather than just process liveness. Test runs require no unauthorized identifiers and negative access probes fail as intended. It confirms whether the targeted action succeeded: use approved synthetic or appropriately deidentified test data with controlled production smoke verification.
Incorrect Answers
A:That acceptance test is directed at a different validation testing situation: An engineering group duplicates a validation helper across six notebooks, and two copies enforce outdated threshold values. Here the actual success criterion is test runs require no unauthorized identifiers and negative access probes fail as intended.
B:That acceptance test is directed at a different validation testing situation: A feature-refresh task writes rows with event time in local timezone while the training join assumes UTC; unit tests of each component pass. Here the actual success criterion is test runs require no unauthorized identifiers and negative access probes fail as intended.
C:That acceptance test is directed at a different validation testing situation: A date feature function treats all incoming timestamp strings as UTC even though some include explicit offsets and a unit test uses only Z-suffixed inputs. Here the actual success criterion is test runs require no unauthorized identifiers and negative access probes fail as intended.
Question 878
The engineering team encounters the following issue: A candidate fraud model has near-identical average error to the champion, but an organization wants to compare rare fraudulent transactions during a brief canary. What investigation would most effectively isolate this behavior?
- Review expected positive-event counts, label delays and confidence intervals for the canary duration.
- Inspect previous model artifact availability, endpoint configuration snapshots and rollback permissions.
- Read the serving endpoint's active model versions and actual traffic percentages after update.
- Compare effective serving traffic split, model identities and alias resolution in the live endpoint.
Explanation
Correct Answer
A:This is the most discriminating check for the observed symptoms. Review expected positive-event counts, label delays and confidence intervals for the canary duration. The underlying technical issue is that rare-outcome evaluation requires sufficient exposure, observation time and mature labels before concluding parity.
Incorrect Answers
B:That diagnostic evidence is directed at a different deployment strategies situation: An ML operator promotes a new model version but its automated rollback points to a model that was deleted from the endpoint configuration. For the stated incident, the decisive observation is review expected positive-event counts, label delays and confidence intervals for the canary duration.
C:That diagnostic evidence is directed at a different deployment strategies situation: A Databricks Model Serving endpoint hosts champion and candidate versions, but a release config accidentally routes 100% of traffic to the candidate instead of 10%. For the stated incident, the decisive observation is review expected positive-event counts, label delays and confidence intervals for the canary duration.
D:That diagnostic evidence is directed at a different deployment strategies situation: During a controlled release, the candidate model's P95 latency triples and engineers want to change the registered model alias without adjusting endpoint routing. For the stated incident, the decisive observation is review expected positive-event counts, label delays and confidence intervals for the canary duration.
Question 879
Review this production model incident: A travel site serves 30,000 bookings per minute and wants to evaluate a new fraud model while limiting customer impact during its first hour. Which mitigation resolves the defect rather than concealing it?
- Send a small approved percentage of eligible live requests to the candidate and expand only after defined gates pass.
- Add a version-compatible request adapter or coordinate client migration with explicit acceptance gates.
- Preserve the approved model artifact and rehearse supported traffic restoration before the release.
- Gate release completion on ready healthy serving entities and validated predictions.
Explanation
Correct Answer
A:This corrects the relevant failure boundary instead of merely hiding the symptom. Send a small approved percentage of eligible live requests to the candidate and expand only after defined gates pass. The issue matters because a full immediate cutover could expose every booking to an untested model defect.
Incorrect Answers
B:That corrective action is directed at a different deployment strategies situation: A rollout updates the serving model but a renamed input column breaks requests from an older mobile client. The relevant corrective action here is send a small approved percentage of eligible live requests to the candidate and expand only after defined gates pass.
C:That corrective action is directed at a different deployment strategies situation: An ML operator promotes a new model version but its automated rollback points to a model that was deleted from the endpoint configuration. The relevant corrective action here is send a small approved percentage of eligible live requests to the candidate and expand only after defined gates pass.
D:That corrective action is directed at a different deployment strategies situation: An endpoint update request succeeds but the newly configured model version is still loading; monitoring marks the release complete immediately. The relevant corrective action here is send a small approved percentage of eligible live requests to the candidate and expand only after defined gates pass.
Question 880
A production ML review contains this situation: A fairness reviewer needs the difference in false-negative rates between two approved cohorts, but the monitor computes an aggregate without cohort labels. What production-safe change best matches the evidence?
- Evaluate performance only on sufficiently matured matched predictions and report pending labels separately.
- Configure a documented threshold with the responsible on-call team and a tested delivery route.
- Calculate slice-specific error rates and a documented derived disparity metric with privacy safeguards.
- Retain a stable reference comparison in addition to short-term change detection.
Explanation
Correct Answer
C:This corrects the relevant failure boundary instead of merely hiding the symptom. Calculate slice-specific error rates and a documented derived disparity metric with privacy safeguards. The issue matters because a material disparity is hidden inside the overall average model performance.
Incorrect Answers
A:That corrective action is directed at a different drift detection and lakehouse monitoring situation: A fraud model predicts instantly but labels arrive 21 days later; an inference dashboard treats unlabelled transactions as negative examples. The relevant corrective action here is calculate slice-specific error rates and a documented derived disparity metric with privacy safeguards.
B:That corrective action is directed at a different drift detection and lakehouse monitoring situation: A fraud-feature drift monitor crosses its threshold but the alert has no assigned recipient and no incident ownership. The relevant corrective action here is calculate slice-specific error rates and a documented derived disparity metric with privacy safeguards.
D:That corrective action is directed at a different drift detection and lakehouse monitoring situation: A streaming payments monitor always compares today with the previous hour, and a slow week-long distribution drift escapes because adjacent hours differ only slightly. The relevant corrective action here is calculate slice-specific error rates and a documented derived disparity metric with privacy safeguards.