CERTIFIED-MACHINE-LEARNING-PROFESSIONAL Question #55: Real Exam Question with Answer & Explanation

Question

Which of the following is a probable response to identifying drift in a machine learning application?

Options

• ANone of these responses
• BRetraining and deploying a model on more recent data
• CAll of these responses
• DRebuilding the machine learning application with a new label variable
• ESunsetting the machine learning application

Explanation

There appears to be an issue with the stated correct answer here. Based on standard machine learning knowledge, C ("All of these responses") is the more defensible answer - not A.

Here's why each option is actually a valid probable response to drift:

• B (Retraining on more recent data): This is the most common and well-documented response to drift. When input distributions or label relationships shift over time, refreshing the model on current data directly addresses the problem.
• D (Rebuilding with a new label variable): When concept drift reveals the original label no longer captures the target phenomenon, redefining the prediction target is a legitimate (if more drastic) response.
• E (Sunsetting the application): If drift is severe and no retraining strategy recovers acceptable performance, retiring the model is a valid business decision.

If A ("None of these") were truly correct, the question would be claiming that retraining, relabeling, and sunsetting are all improbable responses to drift - which contradicts both academic literature and industry practice.

Memory tip: Think of drift responses on a severity spectrum - retrain first, relabel if deeper issues exist, sunset if nothing works. All three sit on the same spectrum, which is why "all of the above" is the strongest answer.

Recommendation: Double-check the answer key source. This looks like a possible key error - C is the answer most ML curricula would accept.

No community discussion yet for this question.