PMLE Monitoring ML Solutions Practice Question
A team deployed a model to a Vertex AI Endpoint and enabled Vertex AI Model Monitoring for skew detection. They notice that training-serving skew metrics are only produced when the endpoint receives traffic, but they want to ensure the skew is computed correctly even during periods of low traffic. Which configuration should they adjust to ensure skew detection remains statistically valid without generating excessive false positives?
⚠ Common exam trap
The trap here is assuming that increasing sampling rate or frequency will fix low-traffic skew detection, when the real issue is insufficient samples per window.
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
✓
Configure a longer monitoring window (e.g., daily) and set a minimum sample size for the monitoring job.
Training-serving skew detection in Vertex AI Model Monitoring relies on comparing distributions of serving data to the training baseline. With low traffic, short windows yield too few samples, making skew estimates noisy. Using a longer window and enforcing a minimum sample size ensures enough data for statistically meaningful comparisons, reducing false alerts while maintaining detection capability.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Enable explainable AI on the endpoint and use feature attributions to detect skew.
Why it's wrong here
Explainable AI provides feature attributions for individual predictions, which can help understand model behavior, but it does not compute training-serving skew metrics or address statistical validity. Skew detection compares feature distributions between training and serving data; feature attributions are not a substitute. This option does not solve the low-traffic skew computation problem.
- ✗
Adjust the drift threshold to a higher value and reduce the sampling rate to 0.1.
Why it's wrong here
Raising the threshold and lowering sampling reduces sensitivity and data volume, which may suppress alerts but does not make the skew computation statistically valid. With few samples, the estimated skew remains unreliable. This configuration trades off detection capability for fewer alerts, but it does not ensure correct skew computation during low traffic periods.
- ✓
Configure a longer monitoring window (e.g., daily) and set a minimum sample size for the monitoring job.
Why this is correct
Vertex AI Model Monitoring allows you to set the monitoring window and a minimum number of samples required before computing skew. By using a longer window, you accumulate more instances, improving statistical power. Setting a minimum sample size prevents the job from running on insufficient data, thus reducing false positives and ensuring valid skew detection even with low traffic.
- ✗
Increase the monitoring frequency to every 10 minutes and set the sampling rate to 1.0.
Why it's wrong here
Increasing frequency and sampling to 1.0 may produce more data points, but with low traffic, each window still contains few samples, leading to high variance and false positives. The core issue is insufficient sample size per window, not the sampling rate or frequency. This approach does not address the statistical validity requirement and can increase cost without solving the problem.
Go deeper
Related to this question
About these practice questions
One of 775 original PMLE practice questions on Courseiva, each with a full explanation and wrong-answer analysis — not exam dumps or protected exam content. Learn why practice questions differ from exam dumps →
JA
Written and reviewed by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
Last reviewed September 2026 · checked against the official Google Cloud exam blueprint
This PMLE practice question is part of Courseiva's free Google Cloud 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 PMLE exam.