PMLE Monitoring ML Solutions Practice Question

You manage a Vertex AI Model Monitoring job on an Endpoint that serves an image classification model. The monitoring job reports feature skew for the input feature 'brightness' but no prediction drift. You want to determine whether the skew is caused by a change in the distribution of incoming images compared to the training data. Which monitoring configuration should you inspect first?

⚠ Common exam trap

A common mix-up: candidates confuse feature skew with prediction drift and looking at output-related thresholds or sampling settings instead of the training baseline.

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

✓

The training dataset specified in the monitoring job's skew configuration.

Feature skew in Vertex AI Model Monitoring is calculated by comparing the distribution of live prediction inputs to the training data distribution. When skew is flagged, the first step is to verify that the training dataset used as the baseline is correct and representative. If the baseline is outdated or mismatched, the skew alert may be misleading. Inspecting the training dataset configuration is therefore the correct initial diagnostic step.

Answer analysis

Option-by-option breakdown

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

✗
The prediction drift configuration's default threshold for the model's output.
Why it's wrong here
Prediction drift monitors changes in the model's output distribution over time, not input feature skew. Since the job reported no prediction drift, adjusting or inspecting this threshold will not explain the input feature skew. The skew is specifically about input features compared to training data, so this configuration is irrelevant to the reported issue.
✗
The endpoint's traffic split configuration.
Why it's wrong here
Traffic split determines how requests are routed between different model versions on an endpoint. It does not influence the computation of feature skew, which compares input features to training data. Changing the traffic split would not resolve or explain why 'brightness' shows skew, making this an unrelated configuration.
✓
The training dataset specified in the monitoring job's skew configuration.
Why this is correct
Feature skew compares the live prediction input distribution to the training data distribution. Inspecting the training dataset baseline reveals whether the baseline itself has shifted or is unrepresentative, which directly explains the reported skew. Without verifying the baseline, you cannot determine if the skew is genuine or an artifact of a stale or incorrect training set.
✗
The sampling rate set in the monitoring job.
Why it's wrong here
Sampling rate controls what fraction of incoming requests are analyzed for monitoring. While too low a sampling rate could affect statistical confidence, it does not cause systematic skew; it only affects the precision of the estimate. The skew is computed from the sampled data against the training baseline, so the training baseline is the primary suspect.

Go deeper

PMLE Monitoring ML Solutions Practice Question

Why each option matters

Option-by-option breakdown

Related to this question