Writing

Decision tree

I trained a scikit-learn decision tree on the classic Iris dataset, tuned its depth with cross-validation, and pulled apart the maths behind every step: Gini impurity, information gain, and feature importance.