Associate Data Practitioner · Free Practice Question Medium
Question 7
A marketing analyst needs to predict which customers are most likely to respond to an upcoming promotion based on purchase history and demographic data. Which BigQuery ML model type would be most appropriate?
-
A
BOOSTED_TREE_CLASSIFIER
-
B
ARIMA_PLUS
-
C
KMEANS
-
D
MATRIX_FACTORIZATION
Reveal correct answer
Correct answer: A
Explanation
Option A is CORRECT. BOOSTED_TREE_CLASSIFIER is ideal for prediction tasks with tabular data like customer purchase history and demographics. It can capture complex, non-linear relationships between features and the target variable (promotion response), handle mixed data types commonly found in customer data, and typically provides higher accuracy than simpler models for this type of classification problem. Boosted trees also offer good interpretability through feature importance, helping marketers understand which factors drive response.
Option B is INCORRECT. ARIMA_PLUS is designed for time-series forecasting of numerical values over time, not for binary classification of whether customers will respond to a promotion. It's suited for predicting metrics like future sales volumes, not for customer-level response prediction.
Option C is INCORRECT. KMEANS is an unsupervised clustering algorithm that groups similar customers together but doesn't predict specific outcomes like promotion response. While clustering could identify customer segments with different propensities, it doesn't directly provide the probability of response that a classifier would.
Option D is INCORRECT. MATRIX_FACTORIZATION is designed for recommendation systems, typically for predicting user-item interactions or ratings. While it could be used for product recommendations, it's not suitable for predicting binary outcomes like promotion response based on customer attributes.
Discussion
Think the marked answer is wrong, or have a better explanation? Share it below — comments appear after review.
