Question 7
Domain 1 — AI Concepts, Terminology, and Use CasesAn insurance company implements unsupervised learning to identify customer segments for targeted marketing campaigns. The data science team has transaction history, demographic information, and interaction patterns but no predefined customer categories. Which algorithm would be most appropriate for this use case? (Select one!)
Correct answer: C
Explanation
K-means clustering fits unsupervised learning because it groups unlabeled data into clusters based on similarity. With transaction history, demographic information, and interaction patterns but “no predefined customer categories,” the goal is to “identify customer segments,” which is exactly what clustering does for targeted marketing.
Why each option is right or wrong
A. Logistic regression to classify customers into predetermined risk categories
B. Linear regression to forecast future premium amounts per customer
C. K-means clustering to group customers with similar characteristics
K-means is the standard unsupervised method for partitioning unlabeled observations into a fixed number of groups by minimizing within-cluster variance, which matches the absence of predefined customer labels here. In practice, it is appropriate when the objective is segmentation from numeric features such as transaction history, demographics, and interaction metrics, because it assigns each customer to the nearest centroid and produces distinct marketing segments without any target variable.
D. Random forest to predict customer lifetime value from historical patterns