You will train a BigQuery ML model with SQL preprocessing but will run predictions on raw, untransformed input. How should you structure training and prediction to avoid data skew at prediction time?
Choose an answer
Tap an option to check your answer.
Correct answer: When creating the model, use BigQuery's TRANSFORM clause to define preprocessing steps. At prediction time, use BigQuery's ML.EVALUATE clause without specifying any transformations on the raw input data..
Why this is the answer
The TRANSFORM clause in BigQuery ML models automatically stores the preprocessing steps applied during training. When ML.PREDICT or ML.EVALUATE is called on new, raw data, BigQuery ML applies these stored transformations internally before generating predictions, ensuring consistency between training and prediction. This prevents data skew, where the model sees different data distributions during training and inference. The other options either re-apply transformations manually, leading to potential inconsistencies, or use external tools like Dataflow, which doesn't leverage BigQuery ML's built-in TRANSFORM functionality for automatic, consistent preprocessing during prediction.
Pass your exam — without the endless answer hunt
Get every verified question and explanation for this exam in one place, and save hours of prep. 1,000+ certifications · 20+ languages · free to start.
Pass your exam faster → No card needed