cert
cert-1
cert-2

Pass Databricks Certified Machine Learning Professional 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
Certified Machine Learning Professional Exam - Verified By Experts
Certified Machine Learning Professional Premium File

Certified Machine Learning Professional Premium File

$69.99
$76.99
  • Premium File 216 Questions & Answers. Last Update: Nov 12, 2026

Whats Included:

  • Latest Questions
  • 100% Accurate Answers
  • Fast Exam Updates
 
$76.99
$69.99
accept 12 downloads in the last 7 days
block-screenshots
Databricks Certified Machine Learning Professional Practice Test 21

Question 200

In an enterprise ML workflow, the reported issue is: A team switches a SparkML training stage to Ray without verifying that its chosen estimator supports Ray's distributed fit contracts. Before modifying the pipeline, which evidence should be collected first?

  1. Inspect runtime GPU memory per worker alongside Ray trial resource requests.
  2. Inspect the estimator's documented execution support and task resource requirements.
  3. Profile model footprint, example throughput and cross-node parameter synchronization cost.
  4. Inspect concurrent log timestamps and artifact overwrite patterns by MLflow run ID.

Explanation

Correct Answer

B:

This is the most discriminating check for the observed symptoms. Inspect the estimator's documented execution support and task resource requirements. The underlying technical issue is that selecting Ray over Spark must be based on the actual training library and distributed API, not the existence of a cluster.

Incorrect Answers

A:

That diagnostic evidence 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. For the stated incident, the decisive observation is inspect the estimator's documented execution support and task resource requirements.

C:

That diagnostic evidence is directed at a different scaling and tuning situation: A tabular classifier fits comfortably on one GPU and 100 million independent examples are available, but a team proposes splitting the small model's layers across hosts. For the stated incident, the decisive observation is inspect the estimator's documented execution support and task resource requirements.

D:

That diagnostic evidence is directed at a different scaling and tuning situation: An Optuna optimization reuses the same MLflow run ID for concurrent trials, causing overlapping parameter values and artifacts. For the stated incident, the decisive observation is inspect the estimator's documented execution support and task resource requirements.

Question 201

A production ML review contains this situation: A real-time client retries every 429 rate-limit error immediately in an unbounded loop and multiplies load on the serving endpoint. What production-safe change best matches the evidence?

  1. Route changes through a single approved configuration source and verify the active endpoint after each change.
  2. Implement bounded retry and jitter with explicit failure handling for capacity exhaustion.
  3. Remove embedded secrets and configure an approved scoped secret retrieval mechanism for the serving process.
  4. Log the tokenizer as a referenced MLflow artifact and load it from the model context in the PyFunc implementation.

Explanation

Correct Answer

B:

This corrects the relevant failure boundary instead of merely hiding the symptom. Implement bounded retry and jitter with explicit failure handling for capacity exhaustion. The issue matters because the client amplifies throttling and creates a self-sustaining inference outage.

Incorrect Answers

A:

That corrective action is directed at a different custom model serving situation: A team edits a model endpoint through the UI while its CI pipeline concurrently applies an older configuration and reverts the intended traffic split. The relevant corrective action here is implement bounded retry and jitter with explicit failure handling for capacity exhaustion.

C:

That corrective action is directed at a different custom model serving situation: An engineer logs a customer-specific API token inside the pickled PythonModel object before registering it to Unity Catalog. The relevant corrective action here is implement bounded retry and jitter with explicit failure handling for capacity exhaustion.

D:

That corrective action is directed at a different custom model serving situation: A custom PyFunc estimator loads a tokenizer file from an engineer's laptop path when called in a serving container. The relevant corrective action here is implement bounded retry and jitter with explicit failure handling for capacity exhaustion.

Question 202

The engineering team encounters the following issue: A mobile app requires thousands of concurrent low-latency scores with online features, but a team plans to query an offline historical feature Delta table for every request. Which experiment most directly demonstrates a successful fix?

  1. A held-out review confirms alert volume and true-positive capture at the frozen threshold.
  2. The full fitted pipeline completes on a small test DataFrame and exposes the planned input and output columns.
  3. Both known and unknown fixture categories yield the documented safe scoring behavior.
  4. Load tests meet prediction latency and feature freshness objectives at planned peak traffic.

Explanation

Correct Answer

D:

This verification measures the required result rather than just process liveness. Load tests meet prediction latency and feature freshness objectives at planned peak traffic. It confirms whether the targeted action succeeded: publish approved features to a supported low-latency online feature serving path and test endpoint integration.

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 load tests meet prediction latency and feature freshness objectives at planned peak traffic.

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 load tests meet prediction latency and feature freshness objectives at planned peak traffic.

C:

That acceptance test is directed at a different using spark ml situation: An inference pipeline sends new categorical labels unseen during training to StringIndexerModel with error handling configured to reject unknown categories. Here the actual success criterion is load tests meet prediction latency and feature freshness objectives at planned peak traffic.

Question 203

A production ML review contains this situation: A global model predicts late deliveries accurately overall, but all mistakes occur at one small coastal depot and the team never slices its results by site. Which check shows that this particular problem has been resolved?

  1. A cohort with large cumulative distance from its training baseline raises a distinct signal.
  2. A fixture with coast-only errors raises its depot-specific metric without changing unrelated depots.
  3. A privacy scan confirms only approved aggregate values appear in the shared metric output.
  4. Metrics change only when genuine label observations are joined to the correct predictions.

Explanation

Correct Answer

B:

This verification measures the required result rather than just process liveness. A fixture with coast-only errors raises its depot-specific metric without changing unrelated depots. It confirms whether the targeted action succeeded: add an approved site-level performance slice with a minimum-sample threshold.

Incorrect Answers

A:

That acceptance test is directed at a different drift detection and lakehouse monitoring situation: A model trained on last year's customer mix is monitored only against yesterday's already shifted cohort, so its original training-population mismatch is hidden. Here the actual success criterion is a fixture with coast-only errors raises its depot-specific metric without changing unrelated depots.

C:

That acceptance test is directed at a different drift detection and lakehouse monitoring situation: A monitor owner writes a custom metric that logs unaggregated account-level health data in a broadly shared results table. Here the actual success criterion is a fixture with coast-only errors raises its depot-specific metric without changing unrelated depots.

D:

That acceptance test 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. Here the actual success criterion is a fixture with coast-only errors raises its depot-specific metric without changing unrelated depots.

Question 204

Review this production model incident: A team runs independent GPU-heavy neural trials sequentially, despite having several isolated GPU workers that can fit each model. Which mitigation resolves the defect rather than concealing it?

  1. Increase the training host's memory within budget or change to a supported distributed estimator.
  2. Pin shared validation protocol and data snapshot for the compared Optuna trials.
  3. Eliminate or batch the serialized dependency before purchasing more training replicas.
  4. Configure supported Ray Tune trials with declared GPU and CPU requirements and controlled concurrency.

Explanation

Correct Answer

D:

This corrects the relevant failure boundary instead of merely hiding the symptom. Configure supported Ray Tune trials with declared GPU and CPU requirements and controlled concurrency. The issue matters because available accelerators sit idle while the tuning window expires.

Incorrect Answers

A:

That corrective action 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 relevant corrective action here is configure supported Ray Tune trials with declared GPU and CPU requirements and controlled concurrency.

B:

That corrective action is directed at a different scaling and tuning situation: A hyperparameter team compares two Optuna studies with different train/validation splits and chooses the lower loss as though both metrics were directly comparable. The relevant corrective action here is configure supported Ray Tune trials with declared GPU and CPU requirements and controlled concurrency.

C:

That corrective action is directed at a different scaling and tuning situation: An ML project has a single sequential external feature-service dependency and expects data-parallel workers to increase throughput 20-fold. The relevant corrective action here is configure supported Ray Tune trials with declared GPU and CPU requirements and controlled concurrency.

Question 205

In an enterprise ML workflow, the reported issue is: A feature function maps a null purchase amount to zero without distinguishing missing observations, and a notebook unit test covers only positive amounts. Before modifying the pipeline, which evidence should be collected first?

  1. Inspect function outputs for null, zero and positive synthetic inputs against the documented business rule.
  2. Compare timestamp values after ingestion, feature-table write and training-set as-of join for one known event.
  3. Inspect notebook entrypoints, test invocation conditions and deployment job definitions.
  4. Compare image digests and dependency manifests across unit, integration and deployment stages.

Explanation

Correct Answer

A:

This is the most discriminating check for the observed symptoms. Inspect function outputs for null, zero and positive synthetic inputs against the documented business rule. The underlying technical issue is that unit tests should exercise edge cases and expected output contracts, not just the happy-path fixture.

Incorrect Answers

B:

That diagnostic evidence 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. For the stated incident, the decisive observation is inspect function outputs for null, zero and positive synthetic inputs against the documented business rule.

C:

That diagnostic evidence is directed at a different validation testing situation: An ML project stores test files directly in each production notebook and execution of a deployed notebook inadvertently reruns expensive tests. For the stated incident, the decisive observation is inspect function outputs for null, zero and positive synthetic inputs against the documented business rule.

D:

That diagnostic evidence is directed at a different validation testing situation: A release pipeline runs development unit tests against a different dependency image from the one deployed to production. For the stated incident, the decisive observation is inspect function outputs for null, zero and positive synthetic inputs against the documented business rule.

Question 206

Review this production model incident: A test suite mocks every external interface so aggressively that even an incompatible feature table schema appears valid. Which failure risk follows most closely from the unresolved behavior?

  1. A simple deterministic transform bug reaches users before it is caught by a heavy live workflow.
  2. Invalid standardized values contaminate model input vectors or crash scoring.
  3. The system passes isolated tests and fails immediately after deployment.
  4. A model that looks accurate offline cannot be called successfully by production clients.

Explanation

Correct Answer

C:

This consequence follows from the specific unresolved condition. The system passes isolated tests and fails immediately after deployment. The causal constraint is: Mocks are appropriate for isolated logic but cannot prove real integration contracts.

Incorrect Answers

A:

That direct consequence is directed at a different validation testing situation: A developer changes a feature transform in a notebook and directly uses production data as the only test case. The more immediate risk here is the system passes isolated tests and fails immediately after deployment.

B:

That direct consequence is directed at a different validation testing situation: A preprocessing function divides by a standard deviation and returns infinity for a constant numeric column. The more immediate risk here is the system passes isolated tests and fails immediately after deployment.

D:

That direct consequence is directed at a different validation testing situation: A churn feature job succeeds and a classifier achieves acceptable F1, but the deployed endpoint returns HTTP 400 because request column names differ. The more immediate risk here is the system passes isolated tests and fails immediately after deployment.

Question 207

Review this production model incident: An automated retraining job triggers whenever a numerical feature mean shifts by 0.1, regardless of normal seasonal changes or sample size. Which failure risk follows most closely from the unresolved behavior?

  1. Normal weekly seasonality launches wasteful retraining runs and noisy alerts.
  2. A statistically better average model causes larger losses for the protected operational segment.
  3. Random sampling noise causes frequent unnecessary production model swaps.
  4. The winning model cannot serve requests on time or within allocated capacity.

Explanation

Correct Answer

A:

This consequence follows from the specific unresolved condition. Normal weekly seasonality launches wasteful retraining runs and noisy alerts. The causal constraint is: Drift-triggered retraining needs statistically and operationally justified thresholds, not arbitrary raw metric movement.

Incorrect Answers

B:

That direct consequence is directed at a different automated retraining situation: A new fraud classifier improves overall F1 but doubles false negatives for high-value payment accounts, and an automated policy chooses it on F1 alone. The more immediate risk here is normal weekly seasonality launches wasteful retraining runs and noisy alerts.

C:

That direct consequence is directed at a different automated retraining situation: A new model beats the champion by 0.1 percentage points on a tiny evaluation set, and the team automatically promotes it despite large uncertainty. The more immediate risk here is normal weekly seasonality launches wasteful retraining runs and noisy alerts.

D:

That direct consequence is directed at a different automated retraining situation: An AutoML training job always chooses the model with the lowest training loss, even when the deployed service has a strict memory and P95 latency budget. The more immediate risk here is normal weekly seasonality launches wasteful retraining runs and noisy alerts.

Question 208

In an enterprise ML workflow, the reported issue is: A test engineer can create a model-serving endpoint in a sandbox but the production deployment fails because the automation identity lacks registered-model access. How should the engineer validate the correction?

  1. A clean isolated model-load test resolves the packaged tokenizer and scores a fixture.
  2. A rollback fixture never returns cached candidate scores to champion-only requests.
  3. The versioned model loads from Unity Catalog in a clean environment and reproduces expected outputs.
  4. The scoped automation can deploy the authorized model and cannot access unrelated catalog models.

Explanation

Correct Answer

D:

This verification measures the required result rather than just process liveness. The scoped automation can deploy the authorized model and cannot access unrelated catalog models. It confirms whether the targeted action succeeded: grant only the required production deployment/model access to a dedicated controlled service identity.

Incorrect Answers

A:

That acceptance test is directed at a different custom model serving situation: A custom PyFunc estimator loads a tokenizer file from an engineer's laptop path when called in a serving container. Here the actual success criterion is the scoped automation can deploy the authorized model and cannot access unrelated catalog models.

B:

That acceptance test is directed at a different custom model serving situation: An integration caches predictions without recording the model version and returns yesterday's values after a controlled rollback. Here the actual success criterion is the scoped automation can deploy the authorized model and cannot access unrelated catalog models.

C:

That acceptance test is directed at a different custom model serving situation: A team logs custom MLflow model weights but forgets to include required Python package versions in the model environment. Here the actual success criterion is the scoped automation can deploy the authorized model and cannot access unrelated catalog models.

Question 209

Review this production model incident: A custom PyFunc's predict method assumes every request contains a numeric feature that is optional under the documented inference schema. What end-to-end verification is most persuasive here?

  1. Boundary requests return validated predictions or controlled informative errors according to contract.
  2. A multirow fixture yields correctly parsed predictions and confidences in original request order.
  3. A rollback fixture never returns cached candidate scores to champion-only requests.
  4. An endpoint reaches a ready state and a test prediction succeeds before the pipeline marks the release complete.

Explanation

Correct Answer

A:

This verification measures the required result rather than just process liveness. Boundary requests return validated predictions or controlled informative errors according to contract. It confirms whether the targeted action succeeded: implement explicit schema validation and approved default/missing-value treatment inside the custom model.

Incorrect Answers

B:

That acceptance test is directed at a different custom model serving situation: A serving client calls the REST endpoint expecting one number but the custom PyFunc returns a structured list of prediction and confidence fields. Here the actual success criterion is boundary requests return validated predictions or controlled informative errors according to contract.

C:

That acceptance test is directed at a different custom model serving situation: An integration caches predictions without recording the model version and returns yesterday's values after a controlled rollback. Here the actual success criterion is boundary requests return validated predictions or controlled informative errors according to contract.

D:

That acceptance test is directed at a different custom model serving situation: A Databricks SDK deployment script sends a model-version update but does not wait for the endpoint to reach a usable ready state before announcing success. Here the actual success criterion is boundary requests return validated predictions or controlled informative errors according to contract.

Question 210

A production ML review contains this situation: One enterprise account holds 80% of historical events, and grouped pandas training for that account exceeds executor memory while other accounts finish. What production-safe change best matches the evidence?

  1. Scale worker capacity and partition layout according to measured distributed CPU demand.
  2. Split or redesign only the oversized model workload with an explicitly approved grouping strategy.
  3. Batch or redesign microtrials and choose a proportionate degree of Ray parallelism.
  4. Use supported data parallelism with appropriately batched examples and convergence checks.

Explanation

Correct Answer

B:

This corrects the relevant failure boundary instead of merely hiding the symptom. Split or redesign only the oversized model workload with an explicitly approved grouping strategy. The issue matters because a single skewed group repeatedly kills workers despite healthy median group runtime.

Incorrect Answers

A:

That corrective action is directed at a different scaling and tuning situation: A feature-engineering job is CPU-bound on hundreds of independent daily partitions and a team upgrades only the driver to a much larger machine. The relevant corrective action here is split or redesign only the oversized model workload with an explicitly approved grouping strategy.

C:

That corrective action is directed at a different scaling and tuning situation: A researcher distributes 500 tiny one-second trials across many Ray workers and sees coordination cost exceed compute time. The relevant corrective action here is split or redesign only the oversized model workload with an explicitly approved grouping strategy.

D:

That corrective action is directed at a different scaling and tuning situation: A tabular classifier fits comfortably on one GPU and 100 million independent examples are available, but a team proposes splitting the small model's layers across hosts. The relevant corrective action here is split or redesign only the oversized model workload with an explicitly approved grouping strategy.

Certified Machine Learning Professional Practice Tests