9.1. Locally modeling complex data

Tree-based regression Pros: Fits complex, nonlinear data Cons: Difficult to interpret results Works with: Numeric values, nominal values

In chapter 3, we used decision trees for classification. Decision trees work by successively splitting the data into smaller segments until all of the target variables are the same or until the dataset can no longer be split. Decision trees are a type of greedy algorithm that makes the best choice at a given time without concern for global optimality.