How to Monitor Feature Attribution Drift Cost-Effectively in Vertex AI?

PMLE By islever Editorial Last updated Aug 31, 2026

Short answer

D Create a feature attribution drift monitoring job. Set the sampling rate to 0.1 and the monitoring frequency to weekly.

Question

You work for a hospital. You received approval to collect the necessary patient data, and you trained a Vertex AI tabular AutoML model that calculates patients' risk score for hospital admission. You deployed the model. However, you're concerned that patient demographics might change over time and alter the feature interactions and impact prediction accuracy. You want to be alerted if feature interactions change, and you want to understand the importance of the features for the predictions. You want your alerting approach to minimize cost. What should you do?

Create a feature drift monitoring job. Set the sampling rate to 1 and the monitoring frequency to weekly.
Create a feature drift monitoring job. Set the sampling rate to 0.1 and the monitoring frequency to weekly.
Create a feature attribution drift monitoring job. Set the sampling rate to 1 and the monitoring frequency to weekly.
Create a feature attribution drift monitoring job. Set the sampling rate to 0.1 and the monitoring frequency to weekly. Correct Answer

Key Takeaways

Tests the distinction between raw input distribution changes and shifts in predictive importance, where candidates frequently miss the financial trade-off of full-rate sampling.

This question evaluates your ability to select the appropriate Vertex AI Model Monitoring configuration for detecting predictive shifts while managing expenses. The community consensus confirms that feature attribution drift monitoring with a 0.1 sampling rate optimally balances accuracy tracking with cost efficiency.

Common mistake: Option C is commonly selected because it correctly identifies feature attribution drift but ignores the explicit instruction to minimize costs, making a 1.0 sampling rate financially unsustainable.

Expert Analysis

Why the Answer Is Correct

Feature attribution drift monitoring specifically tracks how the influence of individual features on model predictions evolves, directly addressing concerns about shifting demographic impacts on feature interactions. Configuring a sampling rate of 0.1 reduces computational overhead by analyzing only 10% of traffic, which satisfies the cost-minimization requirement without sacrificing trend visibility. A weekly schedule offers a pragmatic cadence for healthcare datasets that do not require sub-hourly alerting.

Why the Other Options Are Wrong

Options A and B rely on standard feature drift monitoring, which only flags statistical anomalies in input distributions without explaining how those changes alter prediction outcomes or feature relevance. Option C correctly targets attribution drift but applies a sampling rate of 1.0, processing every single prediction request and violating the explicit budget constraint. Standard drift checks cannot capture interaction shifts or provide the explainability metrics clinicians require for risk assessment validation.

Community Comment Notes

Comment [1] highlights that attribution drift directly measures changes in feature importance, providing an official documentation link for further reading. Comment [3] clearly articulates why a sampling rate of 1.0 generates prohibitive costs in production environments. Comment [2] reinforces that tracking feature importance necessitates attribution monitoring rather than basic distribution checks. These collective insights strongly validate option D as the most balanced solution.

Official Reference

https://cloud.google.com/vertex-ai/docs/model-monitoring/monitor-explainable-ai

This link opens the official vendor documentation in a new tab.

Exam Strategy

Always prioritize explicit operational constraints like cost reduction when evaluating sampling rates; a 0.1 rate is the default industry standard for economical monitoring. Differentiate between data drift and attribution drift by scanning for keywords related to model interpretability versus input statistics.

Community Votes

100%

100% of anonymous learners picked answer D. Votes are pick records left by other test-takers — they are not the verified answer.

Discussion 4 contributions

Compiled from public community threads and reader submissions, lightly edited for clarity. Handles and vote tallies are shown as recorded, and comments reflect the views of individual contributors.

From public community threads 4

Omi_04040 👍 3 Answered D

specifically concerned about changes in feature interactions and their impact on predictions. Feature attribution drift monitoring directly addresses this by tracking how the importance of different features (and their interactions) changes over time. https://cloud.google.com/vertex-ai/docs/model-monitoring/monitor-explainable-ai
AB_C 👍 2 Answered D

Why other options are less suitable: A and B (Feature Drift Monitoring): While basic feature drift monitoring can detect changes in feature distributions, it doesn't directly address your concern about changes in feature interactions and their impact on predictions. C (Sampling Rate of 1): Analyzing 100% of the prediction requests for feature attribution drift can be expensive, especially if you have high traffic.
e821027 👍 2 Answered D

But the interest is also to understand the importance of features for the predictions.
JDpmle2024 👍 1 Answered B

This is feature drift (features are changing) and not feature attribution drift (features are having different effects on the prediction).

Reader comments & corrections

No reader corrections yet — spotted an error or have a note? Share it below.

Practice All PMLE Questions

Access 65 questions with complete answers and detailed explanations.

View Full PMLE Practice Test →

← Back to PMLE Study Guide