Vertex AI Explainable AI and Model Monitoring Configuration

You work at a bank. You need to develop a credit risk model to support loan application decisions. You decide to implement the model by using a neural network in TensorFlow. Due to regulatory requirements, you need to be able to explain the model’s predictions based on its features. When the model is deployed, you also want to monitor the model’s performance over time. You decided to use Vertex AI for both model development and deployment. What should you do?

  1. Use Vertex Explainable AI with the sampled Shapley method, and enable Vertex AI Model Monitoring to check for feature distribution drift. Source Reference Answer
  2. Use Vertex Explainable AI with the sampled Shapley method, and enable Vertex AI Model Monitoring to check for feature distribution skew.
  3. Use Vertex Explainable AI with the XRAI method, and enable Vertex AI Model Monitoring to check for feature distribution drift.
  4. Use Vertex Explainable AI with the XRAI method, and enable Vertex AI Model Monitoring to check for feature distribution skew.

Community Votes

A
100%

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

Community Insight

This question tests the correct pairing of explainability methods for structured data versus images, and the distinction between monitoring for drift over time versus training-serving skew.

For structured credit risk models, use Sampled Shapley for explainability and feature distribution drift for monitoring performance over time. The community agrees that XRAI is strictly for image data and drift is necessary for temporal tracking.

Choosing XRAI methods (Options C and D) is a common error because XRAI is designed for image data, whereas credit risk models use structured tabular data requiring Sampled Shapley.

Community Discussion (3 comments)

b1a8fae 👍 9 Selected: A
Not image -> not XRAI Performance over time -> drift, not skew
fitri001 👍 2 Selected: A
why not the others? B. Feature Distribution Skew: While skew can be relevant, drift is generally a more significant concern for credit risk models. Drift indicates a change in the underlying data distribution, potentially impacting model performance. C & D. XRAI Method: XRAI (Explainable AI for Images) is specifically designed for interpreting image classification models. It wouldn't be the most effective choice for a neural network-based credit risk model working with tabular data.
winston9 👍 2 Selected: A
Explainable AI with the XRAI method is for unstructured, image region analysis, in this case we use structured data for loan approval analysis.

Comments & Corrections

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

Log in to comment, report an error, or add a note about this question.

Submitted for moderation before publishing. Keep it helpful and respectful.

Expert Analysis

Why the Answer Is Correct

Option A is correct because the credit risk model uses structured data, making Sampled Shapley the appropriate explainability method. Additionally, monitoring performance "over time" requires checking for feature distribution drift, which detects changes in the data distribution as time passes.

Why the Other Options Are Wrong

Options C and D are incorrect because XRAI is specifically designed for image-based models to highlight image regions, not tabular data. Option B is incorrect because feature distribution skew compares training data to serving data at a point in time, whereas the requirement is to track changes over time, which necessitates drift detection.

Community Comment Notes

Comments emphasize that XRAI is for "unstructured, image region analysis" and should not be used for loan approval analysis. They also clarify that "performance over time" directly maps to drift, making skew a less suitable choice for this specific requirement.

Official Reference

Exam Strategy

Always identify the data type first: use Sampled Shapley for structured data and XRAI for images. For monitoring, associate "over time" with drift and "training vs. serving" comparison with skew.

Related Analysis

You recently deployed a pipeline in Vertex AI Pipelines that

A-Grade

You need to train an XGBoost model on a small dataset. Your

S-Grade

You are an ML engineer at a retail company. You have built a

S-Grade

Practice All PMLE Questions

Access 65 questions with complete answers and detailed explanations.

View Full PMLE Practice Test →

← Back to PMLE Study Guide