What is the primary goal of an integration test for an ML pipeline?

To test the user interface of the ML application

To measure the performance of the model on a benchmark dataset

To ensure that the different stages of the pipeline work together correctly

To verify that individual functions produce correct outputs

What is a key benefit of placing unit tests in the same notebook as the functions they test?

It reduces the need for version control

It makes tests easier to discover and run alongside the code

It eliminates the need for test data

It ensures tests run in production

When designing an integration test for an ML pipeline, which stages should be covered to ensure the entire pipeline works end-to-end?

Feature engineering, training, evaluation, deployment, and inference

Data ingestion, feature engineering, and training only

A team is building an integration test for their ML pipeline in Databricks. The pipeline includes feature engineering, model training, evaluation, and deployment to a model serving endpoint. They want the integration test to run automatically on every code change. Which approach best satisfies this requirement?

Create a Databricks job that runs the pipeline with a small sample dataset and asserts that the model endpoint returns predictions.

Deploy the model to production and monitor its performance for a week.

Run unit tests on the feature engineering functions and the training code separately.

Use a notebook that only tests the evaluation metrics on historical data.

A team is deciding whether to put their unit tests in the same notebook as the functions or in a separate notebook. What is a key challenge of putting tests in the same notebook?

It makes the notebook longer and harder to read.

It makes the code more portable.

It eliminates the need for integration tests.

DatabricksCertified Machine Learning Professional

Domain 2Objective 2

Validation Testing MACHINE-LEARNING-PROFESSIONAL Practice Questions (Page 3)

Part of the ML Ops domain, which accounts for 44% of the MACHINE-LEARNING-PROFESSIONAL exam. Databricks does not publish an official question count, but from its 120-minute exam (~50–80 total, ~22–35 in this domain), expect 4–7 from this objective — we provide 17 practice questions to prepare you well beyond it. (estimate)

17questions here

4free pages

4concepts

44%of the exam

Questions 11–15

11foundation · easy
What is the primary goal of an integration test for an ML pipeline?
Select an answer first
12foundation · easy
What is a key benefit of placing unit tests in the same notebook as the functions they test?
Select an answer first
13foundation · easy
When designing an integration test for an ML pipeline, which stages should be covered to ensure the entire pipeline works end-to-end?
Select an answer first
14expert · hard
A team is building an integration test for their ML pipeline in Databricks. The pipeline includes feature engineering, model training, evaluation, and deployment to a model serving endpoint. They want the integration test to run automatically on every code change. Which approach best satisfies this requirement?
Select an answer first
15application · medium
A team is deciding whether to put their unit tests in the same notebook as the functions or in a separate notebook. What is a key challenge of putting tests in the same notebook?
Select an answer first

Finished these 5 questions?

Review the revealed explanations, or continue through the curriculum.

Next 5 questions →Back to MACHINE-LEARNING-PROFESSIONAL

Free Basic Practice is a study aid with revealable answers — not a scored exam. Examers.io is independent and not affiliated with or endorsed by Databricks. “MACHINE-LEARNING-PROFESSIONAL” is a trademark of its owner, used for identification only.