In this chapter, you will analyze and apply different base learners in XGBoost. In XGBoost, base learners are the individual models, most commonly trees, that are iterated upon for each boosting round. Along with the default decision tree, which XGBoost defines as
gbtree, additional options for base learners include
dart. Furthermore, XGBoost has its own implementations of random forests as base learners and as tree ensemble algorithms that you will experiment with in this chapter.
By learning how to apply alternative base learners, you will greatly extend your range with XGBoost. You will have the capacity to build many more models and you will learn new approaches to developing linear, tree-based, and random forest machine learning algorithms. The goal of the chapter is to give you proficiency in building XGBoost models with alternative base learners so that you can leverage advanced XGBoost options to find the best possible...