Which of the following data mining models uses both classification and prediction techniques?

• A. Naive Bayes
• B. Decision Trees
• C. K-Means
• D. Neural Networks

Answer: (B)

Decision trees are versatile because they can be built for both predicting a category and estimating a numeric value. When used for classification, each leaf assigns a class label to the instances that reach it. When used for regression (prediction of continuous values), each leaf outputs a numeric estimate, often the average of the target values in that leaf. The same tree-building process—splitting data based on feature values to reduce impurity or error—applies to both tasks, so this model naturally handles classification and prediction within one framework. The other options tend to specialize: Naive Bayes is primarily a classifier, K-Means is clustering without prediction, and neural networks can do both but are typically discussed as separate configurations for classification versus regression, whereas decision trees explicitly cover both within a single approach.