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 20

Question 190

In an enterprise ML workflow, the reported issue is: A fraud model predicts instantly but labels arrive 21 days later; an inference dashboard treats unlabelled transactions as negative examples. Before modifying the pipeline, which evidence should be collected first?

  1. Inspect cohort sizes, allowed sensitive fields and uncertainty around the reported subgroup error rate.
  2. Test the metric with zero, small and normal baseline denominators and inspect SQL null behavior.
  3. Inspect metric reference table identity and original model training-set distribution for the deployed version.
  4. Compare inference IDs and prediction times against actual confirmed label availability.

Explanation

Correct Answer

D:

This is the most discriminating check for the observed symptoms. Compare inference IDs and prediction times against actual confirmed label availability. The underlying technical issue is that true recall and precision require mature observed labels, not fabricated negatives for pending outcomes.

Incorrect Answers

A:

That diagnostic evidence is directed at a different drift detection and lakehouse monitoring situation: A demographic fairness review uses a tiny sensitive subgroup with three labeled examples and reports its observed error as definitive. For the stated incident, the decisive observation is compare inference IDs and prediction times against actual confirmed label availability.

B:

That diagnostic evidence is directed at a different drift detection and lakehouse monitoring situation: A custom drift metric is expressed as today-minus-yesterday percentage, but yesterday's eligible count can be zero. For the stated incident, the decisive observation is compare inference IDs and prediction times against actual confirmed label availability.

C:

That diagnostic evidence 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. For the stated incident, the decisive observation is compare inference IDs and prediction times against actual confirmed label availability.

Question 191

The engineering team encounters the following issue: A scoring endpoint quietly moves from model v7 to v8 during a rollout but an inference trend aggregates both versions in one error curve. Which experiment most directly demonstrates a successful fix?

  1. A safe synthetic threshold crossing reaches the correct team with the monitored metric and window.
  2. A sustained load test meets the approved P95/P99 objectives rather than just average response time.
  3. A cohort with large cumulative distance from its training baseline raises a distinct signal.
  4. A synthetic v8-only error increase appears in its own model-version metric.

Explanation

Correct Answer

D:

This verification measures the required result rather than just process liveness. A synthetic v8-only error increase appears in its own model-version metric. It confirms whether the targeted action succeeded: compute comparable version-specific performance trends and retain traffic-share context.

Incorrect Answers

A:

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 synthetic v8-only error increase appears in its own model-version metric.

B:

That acceptance test is directed at a different drift detection and lakehouse monitoring situation: A custom model endpoint meets mean response time of 75 milliseconds but P99 latency reaches 1.8 seconds during peak load. Here the actual success criterion is a synthetic v8-only error increase appears in its own model-version metric.

C:

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 synthetic v8-only error increase appears in its own model-version metric.

Question 192

In an enterprise ML workflow, the reported issue is: A forecasting model has a few catastrophic numeric errors hidden by a low median absolute error and operations care about large misses. What technical distinction is essential to this investigation?

  1. Hyperparameter selection uses validation evidence; the final unbiased estimate requires an independent untouched test split.
  2. StringIndexer is an Estimator that must fit its category mapping before its resulting model can transform records.
  3. RMSE squares residuals and highlights large errors, while median error alone may obscure expensive outliers.
  4. A SparkML Pipeline requires transformer stages in an order that produces the input types expected by later stages.

Explanation

Correct Answer

C:

The technical constraint is specific to the reported failure. RMSE squares residuals and highlights large errors, while median error alone may obscure expensive outliers. To test it, compare absolute and squared residual distributions and inspect the high-error tail on holdout data.

Incorrect Answers

A:

That underlying principle is directed at a different using spark ml situation: After choosing the best parameter map, a data scientist reports the same cross-validation score as an independent final test result without any untouched data. The current incident instead turns on rMSE squares residuals and highlights large errors, while median error alone may obscure expensive outliers.

B:

That underlying principle is directed at a different using spark ml situation: A new Spark developer calls transform on an unfitted StringIndexer instance and expects learned category labels to appear automatically. The current incident instead turns on rMSE squares residuals and highlights large errors, while median error alone may obscure expensive outliers.

D:

That underlying principle is directed at a different using spark ml situation: An IoT classifier passes raw text categories directly into VectorAssembler ahead of StringIndexer and OneHotEncoder, generating incompatible vector inputs. The current incident instead turns on rMSE squares residuals and highlights large errors, while median error alone may obscure expensive outliers.

Question 193

Review this production model incident: An analyst wants to inspect today’s static customer table for missing values and profile changes but chooses an inference monitor requiring model IDs and predictions. What is the central engineering constraint behind this behavior?

  1. A snapshot profile suits a general current table when no prediction or event-time model-monitoring contract exists.
  2. Missing-value changes can be a quality or distribution signal and should not be silently discarded from monitored populations.
  3. A fixed baseline must be versioned and immutable for the intended comparison period.
  4. Time-series profile configuration needs a valid temporal column with reliable timestamp semantics.

Explanation

Correct Answer

A:

The technical constraint is specific to the reported failure. A snapshot profile suits a general current table when no prediction or event-time model-monitoring contract exists. To test it, inspect the table schema for presence of model, prediction and timestamp columns versus the requested snapshot statistics.

Incorrect Answers

B:

That underlying principle is directed at a different drift detection and lakehouse monitoring situation: A nightly scoring feed drops all null-feature rows before calculating drift, making missingness jump invisible. The current incident instead turns on a snapshot profile suits a general current table when no prediction or event-time model-monitoring contract exists.

C:

That underlying principle is directed at a different drift detection and lakehouse monitoring situation: A baseline table is overwritten every night with the current day's data before a drift calculation, making the comparison always look unchanged. The current incident instead turns on a snapshot profile suits a general current table when no prediction or event-time model-monitoring contract exists.

D:

That underlying principle is directed at a different drift detection and lakehouse monitoring situation: A time-series monitor uses a string-formatted date that cannot be reliably converted to timestamps and silently groups most rows into an unknown bucket. The current incident instead turns on a snapshot profile suits a general current table when no prediction or event-time model-monitoring contract exists.

Question 194

A production ML review contains this situation: An image model needs 90 GB accelerator memory for one training replica while each available GPU has 24 GB and engineers propose only adding more identical nodes. Which statement best explains the failure mode?

  1. Synchronized distributed training requires a coherent parameter-update protocol and checkpoint versioning.
  2. Horizontal replica count alone does not combine GPU memory for one model without a supported parallelization strategy.
  3. Separate Spark tasks do not safely share or synchronously mutate one Python model instance across workers.
  4. Selecting Ray over Spark must be based on the actual training library and distributed API, not the existence of a cluster.

Explanation

Correct Answer

B:

The technical constraint is specific to the reported failure. Horizontal replica count alone does not combine GPU memory for one model without a supported parallelization strategy. To test it, measure per-replica device memory needs and the framework's actual sharding capability.

Incorrect Answers

A:

That underlying principle is directed at a different scaling and tuning situation: A training team averages gradients from multiple workers but lets each worker use a different untracked version of the model. The current incident instead turns on horizontal replica count alone does not combine GPU memory for one model without a supported parallelization strategy.

C:

That underlying principle is directed at a different scaling and tuning situation: An ML engineer tries to parallelize a stateful neural-training loop by executing independent mapPartitions calls that share one mutable global model in the driver. The current incident instead turns on horizontal replica count alone does not combine GPU memory for one model without a supported parallelization strategy.

D:

That underlying principle is directed at a different scaling and tuning situation: A team switches a SparkML training stage to Ray without verifying that its chosen estimator supports Ray's distributed fit contracts. The current incident instead turns on horizontal replica count alone does not combine GPU memory for one model without a supported parallelization strategy.

Question 195

Consider the following fictional Databricks production case: A churn feature job succeeds and a classifier achieves acceptable F1, but the deployed endpoint returns HTTP 400 because request column names differ. What underlying principle determines the most appropriate interpretation?

  1. Release-stage testing must validate the actual artifact and dependency environment being promoted.
  2. Unit tests should verify reproducible content and stable mappings rather than shape alone.
  3. A full ML integration test should traverse raw feature generation through training, registration, deployment and inference contracts.
  4. Shared reusable validation code can prevent silent divergence between independent notebook copies.

Explanation

Correct Answer

C:

The technical constraint is specific to the reported failure. A full ML integration test should traverse raw feature generation through training, registration, deployment and inference contracts. To test it, trace a synthetic labeled record from feature source through registered model to endpoint request and output.

Incorrect Answers

A:

That underlying principle 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. The current incident instead turns on a full ML integration test should traverse raw feature generation through training, registration, deployment and inference contracts.

B:

That underlying principle is directed at a different validation testing situation: A notebook's label encoder produces a different category order depending on an unordered Python set, but tests assert only output length. The current incident instead turns on a full ML integration test should traverse raw feature generation through training, registration, deployment and inference contracts.

D:

That underlying principle 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. The current incident instead turns on a full ML integration test should traverse raw feature generation through training, registration, deployment and inference contracts.

Question 196

The engineering team encounters the following issue: A team logs a custom Python model that imports a package installed only in the author's notebook, and the serving endpoint fails when loading the artifact. What investigation would most effectively isolate this behavior?

  1. Inspect logged model environment files and serving import errors for the missing package.
  2. Inspect metric computation code and raw values across runs for incompatible scale conventions.
  3. Correlate worker logs and metric histories by run ID to expose simultaneous writes to the same run.
  4. Inspect parent_run_id tags, candidate run IDs and the experiment's tracked search configuration.

Explanation

Correct Answer

A:

This is the most discriminating check for the observed symptoms. Inspect logged model environment files and serving import errors for the missing package. The underlying technical issue is that custom MLflow model packaging must include declared compatible dependencies for the independent serving environment.

Incorrect Answers

B:

That diagnostic evidence is directed at a different advanced mlflow usage situation: A team logs a score labeled accuracy with values between 0 and 100 in some runs and between 0 and 1 in others, then ranks runs automatically. For the stated incident, the decisive observation is inspect logged model environment files and serving import errors for the missing package.

C:

That diagnostic evidence is directed at a different advanced mlflow usage situation: A training orchestrator starts parallel child fits but all workers accidentally log into one active MLflow run and overwrite each other's metrics. For the stated incident, the decisive observation is inspect logged model environment files and serving import errors for the missing package.

D:

That diagnostic evidence is directed at a different advanced mlflow usage situation: An experiment tunes 40 candidates but logs each trial as an unrelated root MLflow run, so reviewers cannot reconstruct which search produced a winner. For the stated incident, the decisive observation is inspect logged model environment files and serving import errors for the missing package.

Question 197

Consider the following fictional Databricks production case: A text pipeline repeatedly fits a Tokenizer stage even though tokenization simply converts strings into token arrays without learning a model. What underlying principle determines the most appropriate interpretation?

  1. Hyperparameter selection uses validation evidence; the final unbiased estimate requires an independent untouched test split.
  2. Small workloads may run more efficiently on one node because Spark startup, scheduling and serialization have fixed costs.
  3. Tokenizer is a Transformer and applies transform directly without a separate learned fit step.
  4. Any fitted preprocessing stage can learn from holdout records and must be fitted within the training boundary.

Explanation

Correct Answer

C:

The technical constraint is specific to the reported failure. Tokenizer is a Transformer and applies transform directly without a separate learned fit step. To test it, inspect the stage class API and whether fitting produces any data-dependent parameters.

Incorrect Answers

A:

That underlying principle is directed at a different using spark ml situation: After choosing the best parameter map, a data scientist reports the same cross-validation score as an independent final test result without any untouched data. The current incident instead turns on tokenizer is a Transformer and applies transform directly without a separate learned fit step.

B:

That underlying principle is directed at a different using spark ml situation: A laboratory runs a linear regression every quarter on only 280 clean observations and proposes a 48-worker Spark cluster to reduce a two-second local fit. The current incident instead turns on tokenizer is a Transformer and applies transform directly without a separate learned fit step.

D:

That underlying principle is directed at a different using spark ml situation: A credit-risk pipeline fits StringIndexer on the entire dataset before splitting train and holdout, and rare holdout labels influence its learned category mapping. The current incident instead turns on tokenizer is a Transformer and applies transform directly without a separate learned fit step.

Question 198

Consider the following fictional Databricks production case: 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. Which observation most directly tests the suspected failure boundary?

  1. Inspect artifact provenance, ownership, version history and permitted access to the proposed model storage.
  2. Compare experiment write rights with registered model and serving endpoint permissions for test identities.
  3. Compare release commit identifiers, approved artifacts and runtime dependency versions across dev, test and prod.
  4. Inventory model run IDs, registered versions, job configuration and deployment references.

Explanation

Correct Answer

C:

This is the most discriminating check for the observed symptoms. Compare release commit identifiers, approved artifacts and runtime dependency versions across dev, test and prod. The underlying technical issue is that a deploy-code strategy promotes a tested, versioned code artifact through environments rather than editing production independently.

Incorrect Answers

A:

That diagnostic evidence is directed at a different model lifecycle management situation: An engineer stores model weights in a personal workspace folder and treats that path as a production artifact registry with audit history. For the stated incident, the decisive observation is compare release commit identifiers, approved artifacts and runtime dependency versions across dev, test and prod.

B:

That diagnostic evidence is directed at a different model lifecycle management situation: A platform team proposes using MLflow Tracking alone to enforce production endpoint authorization, ignoring Unity Catalog and service permissions. For the stated incident, the decisive observation is compare release commit identifiers, approved artifacts and runtime dependency versions across dev, test and prod.

D:

That diagnostic evidence is directed at a different model lifecycle management situation: A team records experiments but has no controlled path from a selected registered model to scheduled scoring jobs. For the stated incident, the decisive observation is compare release commit identifiers, approved artifacts and runtime dependency versions across dev, test and prod.

Question 199

In an enterprise ML workflow, the reported issue is: A research team restarts a long hyperparameter sweep and merges all newly executed trials into a new experiment without connecting them to the original parent. What likely operational or model-quality harm does this issue create?

  1. Online performance degrades because the model receives different kinds of feature values than it learned.
  2. The apparent best F1 cannot be reproduced because its hyperparameters were never captured.
  3. One customer's features can contaminate another customer's live score.
  4. Duplicate or omitted trials distort comparison when two partial histories are treated as independent full searches.

Explanation

Correct Answer

D:

This consequence follows from the specific unresolved condition. Duplicate or omitted trials distort comparison when two partial histories are treated as independent full searches. The causal constraint is: A resumed search needs stable experiment and parent-child provenance across partial reruns to support comparison.

Incorrect Answers

A:

That direct consequence is directed at a different advanced mlflow usage situation: A deployed custom model fetches latest feature values during serving while offline evaluation used event-time snapshots without checking the difference. The more immediate risk here is duplicate or omitted trials distort comparison when two partial histories are treated as independent full searches.

B:

That direct consequence is directed at a different advanced mlflow usage situation: A training notebook reports F1 for several models but only logs the final model's parameters in MLflow, leaving metrics and hyperparameters disconnected. The more immediate risk here is duplicate or omitted trials distort comparison when two partial histories are treated as independent full searches.

C:

That direct consequence is directed at a different advanced mlflow usage situation: A fraud PyFunc uses a global mutable feature cache initialized by the first request and mixes data between subsequent users. The more immediate risk here is duplicate or omitted trials distort comparison when two partial histories are treated as independent full searches.

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.

Certified Machine Learning Professional Practice Tests