Question 1 of 13
The most critical consideration when developing a machine learning model with BigQuery ML is preventing label leakage. This ensures the model learns from appropriate data and doesn't inadvertently incorporate information that would be unavailable during real-world predictions. Careful design of the training and serving pipeline, including feature preparation and evaluation, is essential to align the model with the business objective and avoid biased results. Other options address data protection or automation, but they are secondary to preventing label leakage.
Strong Interview Answer
The model’s accuracy and prevent unintended biases, what is the MOST critical consideration during the feature engineering and model training phases?
What to Listen For
- label leakage
- training pipeline
- feature preparation
- business objective
Caution
Do not restate the multiple-choice stem or ask the learner merely to name a product, rule, or term.