A model is trained using historical customer records where each record already contains the correct outcome, such as “churn” or “not churn.” The…

This scenario is a classic example of supervised learning. The key characteristic is the use of a labeled dataset, where historical data (customer records) includes the correct output or target variable ("churn" or "not churn"). The model learns a mapping function from the input features (customer attributes) to the known output labels. The goal is to train the model on this historical data so it can accurately predict the outcome for new, unseen data (future customers). This process of learning from labeled examples to make predictions is the definition of supervised learning.

B. Unsupervised learning works with unlabeled data to find hidden patterns, such as grouping customers into segments without prior knowledge of those segments.

C. Reinforcement learning involves an agent learning to make decisions by taking actions in an environment to maximize a cumulative reward.

D. Generative learning focuses on creating new data that resembles the training data, not on predicting a specific label from a fixed set of classes.

E. Clustering is a specific technique within unsupervised learning used to group similar data points, which requires unlabeled data.

1. Abu-Mostafa, Y. S., Magdon-Ismail, M., & Lin, H. T. (2012). Learning from Data. AMLBook. Chapter 1, Section 1.1, Page 4. "A credit card company has a wealth of historical data on its customers... This is a supervised learning problem."

2. Stanford University. (2023). CS229: Machine Learning. Course Notes, "Supervised Learning". Defines supervised learning as the task of learning a function that maps an input to an output based on example input-output pairs. The churn prediction problem fits this definition perfectly.