Decision Tree

DecisionTree is a Ruby library that implements decision tree learning with the ID3 information-gain algorithm. It can train models from discrete, continuous, or mixed attribute data. Continuous features are evaluated across possible split points to build threshold-based binary branches. Discrete models classify unique labels and can be rendered for visual inspection. The library supports inconsistent datasets, multiple or symbolic outputs, and fallback values when no branch matches an input. Its Ruleset trainer converts a tree into pruned rules using a held-out portion of the training data. A bagging trainer builds ten rulesets and combines their predictions through voting.