SnowPro Advanced Data ScientistDemo

By Snowflake · Question Mode

QuestionQ81

Train and use machine learning models

A company wants to keep customers involved in its loyalty program by using targeted campaigns for customers who are likely to leave it. The company has historical customer data, including the purchase behavior and marketing interactions of customers who left the program and of current members.

A Data Scientist has been asked to assign every current customer a score representing the likelihood that they will cancel their membership.

What kind of model should be built from this data?

  • A Unsupervised learning: Clustering
  • B Unsupervised learning: Anomaly detection
  • C Supervised learning: Causal inference
  • D Supervised learning: Classification model
Explanation

A supervised classification model can train on historical customer features with a labeled outcome indicating whether each customer left the program, then produce a probability or score for the binary cancellation outcome for current customers. AWS describes binary classification as supervised learning using correctly labeled examples to assign one of two predefined classes.

Learn more

Community Discussion

No comments yet. Be the first to start the discussion!