What if a machine could learn to make decisions by asking a series of simple yes/no questions, just like a doctor diagnosing a patient? That is the core idea behind decision trees — and when we combine many trees, we get some of the most powerful and widely used machine learning methods. In this chapter, you will see how these easy-to-understand models split data into groups, how to stop them from memorising noise, and how ensembles like random forests and gradient boosting turn a collection of weak learners into a strong predictor.