nerdexam
Google

PROFESSIONAL-MACHINE-LEARNING-ENGINEER · Question #171

You work for a magazine distributor and need to build a model that predicts which customers will renew their subscriptions for the upcoming year. Using your company's historical data as your training

The correct answer is B. Use Vertex Explainable AI. Submit each prediction request with the explain' keyword to retrieve. To determine the most predictive customer attribute for each individual prediction from a TensorFlow model deployed on Vertex AI, you should use Vertex Explainable AI.

Submitted by marco_it· Apr 18, 2026Monitoring, optimizing, and maintaining ML solutions

Question

You work for a magazine distributor and need to build a model that predicts which customers will renew their subscriptions for the upcoming year. Using your company's historical data as your training set, you created a TensorFlow model and deployed it to Vertex AI. You need to determine which customer attribute has the most predictive power for each prediction served by the model. What should you do?

Options

  • AStream prediction results to BigQuery. Use BigQuery's CORR(X1, X2) function to calculate the
  • BUse Vertex Explainable AI. Submit each prediction request with the explain' keyword to retrieve
  • CUse Vertex AI Workbench user-managed notebooks to perform a Lasso regression analysis on
  • DUse the What-If tool in Google Cloud to determine how your model will perform when individual

How the community answered

(25 responses)
  • A
    4% (1)
  • B
    76% (19)
  • C
    12% (3)
  • D
    8% (2)

Why each option

To determine the most predictive customer attribute for each individual prediction from a TensorFlow model deployed on Vertex AI, you should use Vertex Explainable AI.

AStream prediction results to BigQuery. Use BigQuery's CORR(X1, X2) function to calculate the

Calculating correlation in BigQuery would provide aggregate relationships between features and outcomes, not per-instance feature importance for individual predictions from a deployed model.

BUse Vertex Explainable AI. Submit each prediction request with the explain' keyword to retrieveCorrect

Vertex Explainable AI provides feature attributions, quantifying how much each input feature contributed to a model's individual prediction, which directly addresses the requirement to determine per-instance predictive power. Submitting prediction requests with the 'explain' keyword retrieves these specific explanations for each prediction.

CUse Vertex AI Workbench user-managed notebooks to perform a Lasso regression analysis on

Performing Lasso regression analysis is a method for feature selection during model development, not for explaining individual predictions from an already deployed model in production.

DUse the What-If tool in Google Cloud to determine how your model will perform when individual

The What-If Tool allows interactive exploration of model behavior by changing feature values, but it's primarily for debugging and 'what-if' scenarios, not for programmatically retrieving feature importance for every prediction served.

Concept tested: Vertex Explainable AI for per-instance feature attributions

Source: https://cloud.google.com/vertex-ai/docs/explainable-ai/overview

Topics

#Vertex Explainable AI#Feature attribution#Model explainability#Google Cloud ML

Community Discussion

No community discussion yet for this question.

Sign up to join the discussion
Full PROFESSIONAL-MACHINE-LEARNING-ENGINEER Practice