PMLE Architecting Low-Code ML Solutions Practice Question

An engineer wants to use BigQuery ML to explain predictions from a trained boosted tree classifier for a specific set of input rows. Which function should they use?

⚠ Common exam trap

PMLE often tests the confusion between global explainability (ML.FEATURE_IMPORTANCE, ML.GLOBAL_EXPLAIN) and local explainability (ML.EXPLAIN_PREDICT) — candidates must know which function provides per-row feature attributions.

Answer choices

Why each option matters

Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.

Correct answer & explanation

✓

ML.EXPLAIN_PREDICT

ML.EXPLAIN_PREDICT is the BigQuery ML function specifically designed to explain predictions from a trained model for a given set of input rows. It returns the prediction along with feature attributions (e.g., Shapley values) that show how each feature contributed to the prediction. This is the correct function for interpreting individual predictions from a boosted tree classifier.

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

✗
ML.EVALUATE
Why it's wrong here
ML.EVALUATE computes aggregate metrics such as accuracy, precision and recall over an entire dataset, not per-row feature attributions. It is tempting because it genuinely assesses model quality, making it correct when validating overall performance rather than explaining why one specific prediction was made.
✗
ML.FEATURE_IMPORTANCE
Why it's wrong here
ML.FEATURE_IMPORTANCE returns global scores for the whole model, not per-row attributions for the specified inputs. It is tempting because it does explain which features drive the model overall, making it the right choice when you need a single global ranking rather than local explanations for individual rows.
✗
ML.PREDICT
Why it's wrong here
ML.PREDICT returns scores or labels, not per-feature attributions, so it cannot explain why a row received its classification. It is the right function for serving a trained model against new data, but explaining predictions requires ML.EXPLAIN_PREDICT, which surfaces the contributing features.
✓
ML.EXPLAIN_PREDICT
Why this is correct
ML.EXPLAIN_PREDICT returns predictions alongside feature attributions for the supplied rows, using explainable AI methods such as Shapley values or integrated gradients, which is exactly the per-row explanation the engineer needs from the boosted tree classifier.

Go deeper

PMLE Architecting Low-Code ML Solutions Practice Question

Why each option matters

Option-by-option breakdown

Related to this question