Decision Tree
A decision tree recursively splits the data space; each internal node branches on a feature threshold and each leaf yields a prediction, offering strong interpretability.
Purpose
To provide an interpretable classification and regression model, the basis of random forests and gradient boosting.
Mechanism
Splits maximize information gain (ID3/C4.5) or minimize Gini impurity (CART). Recursion continues until purity criteria or depth limits are met. Pruning prevents overfitting.
Use cases
- Classification and regression
- Interpretable ML
- Feature selection