Decision Trees and Random Forests from Scratch

Implement Gini impurity and information gain calculations for finding optimal splits.

Implement a function to find the best split for a dataset by iterating through features and values.

Build a recursive function to construct a decision tree with stopping criteria (e.g., max depth, min samples).

Implement bootstrap sampling ('bagging') from scratch.

Build a Random Forest classifier by training multiple decision trees on bootstrapped samples with random feature subsets.

Implement out-of-bag (OOB) error estimation as an alternative to a validation set.