nerdexam
Google

PROFESSIONAL-MACHINE-LEARNING-ENGINEER · Question #236

You work for a telecommunications company. You're building a model to predict which customers may fail to pay their next phone bill. The purpose of this model is to proactively offer at-risk…

The correct answer is D. Define a fairness metric that is represented by accuracy across the sensitive features. Train a. To investigate and mitigate bias in a credit risk model, define fairness metrics across sensitive features and use the BigQuery ML TRANSFORM clause to preprocess these features.

Submitted by omar99· Apr 18, 2026ML model development

Question

You work for a telecommunications company. You're building a model to predict which customers may fail to pay their next phone bill. The purpose of this model is to proactively offer at-risk customers assistance such as service discounts and bill deadline extensions. The data is stored in BigQuery and the predictive features that are available for model training include: - Customer_id - Age - Salary (measured in local currency) - Sex - Average bill value (measured in local currency) - Number of phone calls in the last month (integer) - Average duration of phone calls (measured in minutes) You need to investigate and mitigate potential bias against disadvantaged groups, while preserving model accuracy. What should you do?

Options

  • ADetermine whether there is a meaningful correlation between the sensitive features and the other
  • BTrain a BigQuery ML boosted trees classification model with all features. Use the
  • CTrain a BigQuery ML boosted trees classification model with all features. Use the
  • DDefine a fairness metric that is represented by accuracy across the sensitive features. Train a

How the community answered

(46 responses)
  • A
    7% (3)
  • B
    13% (6)
  • C
    4% (2)
  • D
    76% (35)

Why each option

To investigate and mitigate bias in a credit risk model, define fairness metrics across sensitive features and use the BigQuery ML `TRANSFORM` clause to preprocess these features.

ADetermine whether there is a meaningful correlation between the sensitive features and the other

Determining correlations is an initial step for bias detection but does not provide a strategy for mitigating the identified bias.

BTrain a BigQuery ML boosted trees classification model with all features. Use the

Training with all features and then using SHAP values primarily provides model explainability, not direct bias mitigation, and does not inherently address or reduce bias against disadvantaged groups.

CTrain a BigQuery ML boosted trees classification model with all features. Use the

Similar to SHAP, using LIME primarily provides local model explainability and does not directly mitigate bias in the model's predictions.

DDefine a fairness metric that is represented by accuracy across the sensitive features. Train aCorrect

Defining fairness metrics based on accuracy across sensitive features (e.g., Age, Sex) is crucial for detecting bias. Utilizing the `TRANSFORM` clause in BigQuery ML allows for pre-processing sensitive features, such as bucketing or other transformations, before model training to mitigate their direct influence and reduce bias while potentially preserving overall model accuracy.

Concept tested: Bias detection and mitigation in BigQuery ML

Source: https://cloud.google.com/bigquery-ml/docs/transformations-overview

Topics

#ML Fairness#Bias Mitigation#Fairness Metrics#Responsible AI

Community Discussion

No community discussion yet for this question.

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