PMLE Monitoring ML Solutions Practice Question
A financial institution has deployed a fraud detection model on a Vertex AI Endpoint. The model uses both numerical and categorical features. They have enabled Vertex AI Model Monitoring with training data and configured drift thresholds. After several weeks, they notice that the model's recall has dropped significantly, but the overall prediction distribution remains stable. They suspect that a specific subgroup of transactions is being misclassified. Which approach should they use to identify the subgroup and diagnose the issue?
⚠ Common exam trap
The trap here is assuming that Vertex AI Model Monitoring automatically provides subgroup performance metrics, when in fact it focuses on drift and requires exporting data for deeper analysis.
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
✓
Export the serving data and predictions to BigQuery, then use slice-based evaluation to compare performance metrics across different subgroups.
Exporting serving data and predictions to BigQuery enables slice-based evaluation, which can reveal performance disparities across subgroups. Since Vertex AI Model Monitoring does not provide subgroup-level metrics, this manual analysis is necessary to identify the problematic subgroup and diagnose the recall drop.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Use Vertex AI Model Monitoring's feature attribution analysis to identify which features contribute most to the misclassifications.
Why it's wrong here
Vertex AI Model Monitoring does not provide feature attribution analysis; that is a function of Vertex AI Explainable AI. Model Monitoring focuses on drift detection and does not perform per-prediction attribution. Using Explainable AI could help understand feature importance, but it is not the correct tool for subgroup identification within the monitoring context.
- ✓
Export the serving data and predictions to BigQuery, then use slice-based evaluation to compare performance metrics across different subgroups.
Why this is correct
Exporting serving data and predictions to BigQuery allows you to perform slice-based analysis, such as comparing recall across different subgroups defined by categorical features. This approach can identify which subgroup is experiencing performance degradation. Vertex AI Model Monitoring itself does not provide subgroup-level performance metrics, so this manual analysis is necessary to diagnose the issue.
- ✗
Configure Vertex AI Model Monitoring to monitor the 'transaction_amount' feature with a lower threshold to detect subtle drifts that might affect recall.
Why it's wrong here
Lowering the drift threshold for a single feature may increase sensitivity to changes in that feature, but it does not directly identify which subgroup is misclassified. The issue is not necessarily drift in transaction_amount; it could be a change in the relationship between features and labels for a specific subgroup. This approach does not address the need for subgroup-level performance analysis.
- ✗
Retrain the model immediately using the most recent data to improve recall across all subgroups.
Why it's wrong here
Retraining without diagnosing the cause may not fix the subgroup issue and could degrade performance elsewhere. The problem is specific to a subgroup, so blindly retraining on all recent data might not address the underlying cause, such as a change in fraud patterns for that subgroup. Diagnosis should precede retraining to ensure targeted improvements.
Go deeper
Related to this question
About these practice questions
Courseiva writes every PMLE question from scratch — 775 in total, each with an explanation and a wrong-answer breakdown. None are copied from real exams or dumps. 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.