This project implements Gradient Boosting from scratch, using sklearn for decision trees while handling boosting logic manually. It iteratively trains weak learners to minimize residual errors, improving predictions over multiple iterations. Built with Python, it provides insight into the inner workings of boosting.