Câu 27: SnowPro Advanced Data Scientist DSA-C03

A Data Scientist is developing a two-stage machine learning process for fraud detection. The process needs to first group customers with similar buying behavior into peer groups, and then try to detect abnormal purchase behavior relative to a customer’s peer group. What should the Data Scientist do?

1. A. Use a supervised model to segment customers into peer groups. Then train another supervised model based on the labeled data to perform fraud detection within each peer group.
2. B. Use an unsupervised model for customer segmentation into peer groups. Then use derived examples of normal and abnormal purchases to train a supervised model for fraud detection. — đáp án hiện tại
3. C. Use labeled data to train association rules for customer segmentation into peer groups. Then train an unsupervised regression model to detect customer behavior that deviates from the average behavior in the peer group.
4. D. Use a neural network with two inner layers. The first layer should assign a given customer to a peer group. Then the second layer should perform anomaly detection within the peer group that a customer has been assigned to.