Multiple Choice Flashcards on 02 Databricks Certified Machine Learning Profes...

Question Type

Reader

Answer workspace

Explanation appears after submit

FREE AI FEEDBACK

Why this matters

Review the keyed rationale after answering so the correct AWS service choice and each distractor stay connected.

On demand Read aloud

0 Recent saved answers

Recent Databricks Certified Machine Learning Professional Answers

Active recall loop

Answer, inspect the rationale, then repeat.

Large answer targets and a focused review step keep the current AWS objective at the center of the session.

Answer Review Repeat

Flashcard Practice Databricks Certified Machine Learning Professional Multiple Choice

A bank's approval model has stable global AUC, but auditors ask whether performance has degraded for a protected customer segment. The inference log includes labels, predictions, prediction probabilities, and a Boolean segment indicator. Which approach best supports the audit?

Choose the Best Answer

Use endpoint health metrics because they report protected-segment AUC by default. Use baseline drift only because drift is identical to label-based model performance. Use build logs because fairness metrics are generated when the serving image is created. Use sliced inference quality and fairness-related metrics over time.

Load another random question

More

Question

Choose the Best Answer

How can I help?