Question 36
Domain 5: Monitor, retrain, and manage ML lifecycleYou are using Azure Machine Learning SDK v2 to monitor data drift. You have registered a baseline dataset (dataset1) and a target dataset (dataset2). You want to schedule the drift detection to run every week and track changes in a specific table. Which configuration object should you use to set up this monitoring schedule?
Correct answer: C
Explanation
Azure Machine Learning SDK v2 uses a schedule object to run monitoring jobs on a recurring basis, and the prompt specifies running drift detection "every week." A monitoring schedule is the configuration that ties the baseline and target datasets to a recurring execution, so MonitorSchedule() is the right object to define that weekly drift check.
Why each option is right or wrong
A. DataDriftDetector.run()
B. ml_client.data_drift.create_schedule()
C. MonitorSchedule()
Azure Machine Learning SDK v2 defines recurring monitoring through a schedule resource, and the weekly cadence is expressed by configuring that schedule rather than the drift model itself. In the v2 monitoring API, `MonitorSchedule` is the object used to bind the monitor to a recurrence such as every 7 days, so it is the correct configuration for a weekly run against the specified table.
D. DataQualitySignal()