Decision Tree

Updated 02 Oct 2026

•
What it is
- •
  a predictive model that repeatedly splits observations using feature-based tests and produces predictions at leaves
•
Structure
- •
  root: initial decision node.
- •
  internal nodes: further tests.
- •
  branches: test outcomes.
- •
  leaves: predicted classes, probabilities, or numerical values.
•
Algorithms
- •
  ID3: classification-tree construction using Information Gain.
- •
  C4.5: extends ID3, including gain-ratio-based selection and support for continuous attributes and pruning.
- •
  CART: binary trees for classification or regression.
•
Corrections
- •
  classic ID3 is a classification algorithm; calling ID3 itself a general regression algorithm is inaccurate.
- •
  the practicum’s calculation using information gain alone demonstrates an entropy-based split, not the complete C4.5 selection procedure.
•
Related
- •
  Unavailable reference
- •
  Unavailable reference
- •
  Regression (Data Science)

Linked references 2