A retail company monitors a demand-forecasting model. They want to evaluate model performance separately for each product category and also at a daily granularity. Which approach should they use in Lakehouse Monitoring?

Use feature slicing to group by product category and set the time granularity to daily.

Create separate monitors for each product category.

Use a snapshot table to capture daily performance.

Set the baseline to each product category.

A team monitors a churn model with a Lakehouse Monitor on an inference table. The drift metrics table shows a KS test p-value of 0.02 for 'tenure' and a chi-squared p-value of 0.03 for 'contract_type'. The team has an alert rule that triggers when any drift metric p-value is below 0.05. However, the team is concerned about alert fatigue because the alert fires every week. What should they do?

Disable the alert for 'tenure' because KS tests are unreliable.

Change the baseline to the most recent data to reduce drift.

Lower the alert threshold to p < 0.01 to reduce false positives.

Investigate the root cause of the recurring drift before adjusting the alert.

What is the purpose of feature slicing in Lakehouse Monitoring?

To reduce the amount of data stored in the metrics table

To combine multiple metrics into a single value

To analyze metrics for specific segments of data

To automatically retrain the model for each slice

A model was trained on data from January. In June, the team wants to detect drift by comparing the June data distribution to the January baseline. Which statistical method is most appropriate for a numerical feature like 'income'?

Population Stability Index (PSI)

DatabricksCertified Machine Learning Professional

Domain 2Objective 5

Drift Detection and Lakehouse Monitoring MACHINE-LEARNING-PROFESSIONAL Practice Questions (Page 7)

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 34 practice questions to prepare you well beyond it. (estimate)

34questions here

7free pages

10concepts

44%of the exam

Questions 31–34

31application · medium
A retail company monitors a demand-forecasting model. They want to evaluate model performance separately for each product category and also at a daily granularity. Which approach should they use in Lakehouse Monitoring?
Select an answer first
32expert · hard
A team monitors a churn model with a Lakehouse Monitor on an inference table. The drift metrics table shows a KS test p-value of 0.02 for 'tenure' and a chi-squared p-value of 0.03 for 'contract_type'. The team has an alert rule that triggers when any drift metric p-value is below 0.05. However, the team is concerned about alert fatigue because the alert fires every week. What should they do?
Select an answer first
33foundation · easy
What is the purpose of feature slicing in Lakehouse Monitoring?
Select an answer first
34application · medium
A model was trained on data from January. In June, the team wants to detect drift by comparing the June data distribution to the January baseline. Which statistical method is most appropriate for a numerical feature like 'income'?
Select an answer first

Finished these 4 questions?

Review the revealed explanations, or continue through the curriculum.

No more pagesBack 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.

Drift Detection and Lakehouse Monitoring MACHINE-LEARNING-PROFESSIONAL Practice Questions (Page 7)

Questions 31–34

A retail company monitors a demand-forecasting model. They want to evaluate model performance separately for each product category and also at a daily granularity. Which approach should they use in Lakehouse Monitoring?

What is the purpose of feature slicing in Lakehouse Monitoring?

A model was trained on data from January. In June, the team wants to detect drift by comparing the June data distribution to the January baseline. Which statistical method is most appropriate for a numerical feature like 'income'?

Keep studying — ML Ops