Twenty Questions, Asked Greedily: Decision Trees, Random Forests and Gradient Boosting

Deep End Era. Engineering with Sankalp · Oct 28, 2026 · 55m

Hélène, fresh from a deep learning lab, bets a box of pastries that her neural network will beat actuary-turned-data-scientist Kwame at predicting motor insurance claims. Kwame's afternoon of gradient boosting wins. Part four of The Deep End's Machine Learning series explains why. We grow a tree by ear, meet the survey researchers and statisticians who invented trees, watch Leo Breiman turn instability into strength with bagging and random forests, and follow boosting from a theoretical puzzle about weak learners to a Gödel Prize and real-time face detection. Gradient boosting becomes a round…