This document discusses classification and prediction in data analysis. It defines classification as predicting categorical class labels, such as predicting if a loan applicant is risky or safe. Prediction predicts continuous numeric values, such as predicting how much a customer will spend. The document provides examples of classification, including a bank predicting loan risk and a company predicting computer purchases. It also provides an example of prediction, where a company predicts customer spending. It then discusses how classification works, including building a classifier model from training data and using the model to classify new data. Finally, it discusses decision tree induction for classification and the k-means algorithm.