ChefBoost: A Lightweight Boosted Decision Tree Framework
Sefik Ilkin Serengil · Zenodo (CERN European Organization for Nuclear Research) · 2021
Decision tree based models overwhelmingly over-perform in applied machine learning studies. In this paper, first of all a review decision tree algorithms such as ID3, C4.5, CART, CHAID, Regression Trees and some bagging and boosting methods such as Gradient Boosting, Adaboost and Random Forest have been done and then the description of the developed lightweight boosted decision tree framework - ChefBoost - has been made. Due to its widespread use and intensive choice as a machine learning programming language; Python was selected for the development of framework published also as open source package under MIT license. Moreover, the framework will build decision trees with regular if and else statements as an output. In this way, those statements can be produced and consumed programming language independently.