XGBoost: The Definitive Guide (Part 1)

XGBoost (short for eXtreme Gradient Boosting) is an open-source library that provides an optimized and scalable implementation of gradient boosted decision trees. It incorporates various software and hardware optimization techniques that allow it to deal with huge amounts of data. Originally developed as a research project by Tianqi Chen and Carlos Guestrin in 2016 [1], XGBoost has become the go-to solution for solving supervised learning tasks on structured (tabular) data. It provides state-of-the-art results on many standard regression and classification tasks, and many Kaggle competition winners have used XGBoost as part of their winning solutions. Although significant progress has been made using deep neural networks for tabular data, they are still outperformed by XGBoost and other tree-based models on many standard benchmarks [2, 3]. In addition, XGBoost requires much less tuning than deep models. The main innovations of XGBoost with respect to other gradient boosting algorithms include <a href="https://towardsdatascience.com/xgboost-the-definitive-guide-part-1-cc24d2dcd87a">Read More</a>