Decision and Classification Tree Schematic: Tree-based machine learning methods infer rules for splitting a data set into more homogeneous data sets until a specified number of terminal classes or maximum variance within the terminal classes is reached. The inferred splitting rules can give additional information about the system being studied.