Automated Concept Drift Detection with SageMaker Model Monitor

An e-commerce company uses a machine learning model to predict customer churn. They notice that the model's performance degrades after a major marketing campaign changes customer behavior. Which approach is MOST effective to detect and respond to this type of concept drift?

Quick Answer

The answer is to use SageMaker Model Monitor to track prediction distribution and trigger retraining. This is the most effective approach because concept drift, where the statistical properties of the target variable change due to shifts in customer behavior (like a marketing campaign), directly alters the model’s prediction distribution. SageMaker Model Monitor detects this by continuously comparing live inference data against a baseline, alerting you when the distribution deviates beyond a threshold, and can automatically invoke a retraining pipeline to adapt the model. On the AWS Certified Machine Learning Engineer Associate MLA-C01 exam, this scenario tests your understanding of monitoring strategies versus retraining triggers—a common trap is choosing manual retraining or only tracking data quality (e.g., missing values), which misses the core drift signal. Remember the mnemonic: “Drift in distribution demands detection and retraining.”

⚠ Common exam trap

Many candidates confuse operational scaling (Option E) or periodic evaluation (Option D) with automated drift detection, overlooking that SageMaker Model Monitor provides continuous, distribution-based monitoring and automated retraining triggers.

Answer choices

Why each option matters

Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.

Correct answer & explanation

✓

Use SageMaker Model Monitor to track prediction distribution and trigger retraining.

SageMaker Model Monitor can automatically detect concept drift by tracking the distribution of predictions over time and comparing them against a baseline. When drift is detected, it can trigger a retraining pipeline, enabling the model to adapt to the new customer behavior caused by the marketing campaign without manual intervention.

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

✗
Deploy an A/B test to compare the current model with a baseline.
Why it's wrong here
A/B testing compares model versions but does not inherently detect drift.
✓
Use SageMaker Model Monitor to track prediction distribution and trigger retraining.
Why this is correct
Correct. Model Monitor continuously checks for drift and can initiate automated retraining.
✗
Manually review model accuracy each month.
Why it's wrong here
Manual review is slow and not automated, missing early drift detection.
✗
Set up a weekly batch transform job to compute accuracy against historical data.
Why it's wrong here
Weekly batch is reactive and may not catch drift quickly; also accuracy may not be available without ground truth.
✗
Increase the number of instances for the endpoint.
Why it's wrong here
Scaling instances does not address model degradation.

About these practice questions

Courseiva creates original exam-style practice questions with explanations and wrong-answer analysis. It does not publish real exam questions, exam dumps, or protected exam content. Learn why practice questions differ from exam dumps →

How Courseiva writes practice questions · Editorial policy

Same concept, more angles

1 more way this is tested on MLA-C01

These questions test the same concept from different angles. Work through them to make sure you can recognise it however the exam phrases it.

Variation 1. Refer to the exhibit. A data scientist reviews the CloudWatch Logs from an Amazon SageMaker real-time endpoint. What is the MOST likely root cause of the NaN output?

easy

A.The model weights became corrupted due to a disk write error.
✓ B.The input data contains out-of-range values not seen during training, causing the model to output NaN.
C.The endpoint is overloaded and returning a default NaN response.
D.The model artifact failed to load correctly, resulting in NaN weights.

Why B: The NaN (Not a Number) output from a SageMaker real-time endpoint is most commonly caused by input data containing values outside the range seen during training. This can lead to numerical instability in the model's forward pass, such as division by zero, log of zero, or exponent overflow, which propagates through layers and results in NaN predictions.

Last reviewed: Jul 4, 2026

Question Discussion

Share a tip, memory trick, or ask about the reasoning behind this question. Do not post real exam questions, leaked content, braindumps, or copyrighted exam material. Comments are moderated and may be removed without notice.

Loading comments…

This MLA-C01 practice question is part of Courseiva's free Amazon Web Services certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the MLA-C01 exam.