Summary

In this chapter, you have taken a big leap toward mastering XGBoost by examining decision trees, the primary XGBoost base learners. You built decision tree regressors and classifiers by fine-tuning hyperparameters with GridSearchCV and RandomizedSearchCV. You visualized decision trees and analyzed their errors and accuracy in terms of variance and bias. Furthermore, you learned about an indispensable tool, feature_importances_, which is used to communicate the most important features of your model that is also an attribute of XGBoost.

In the next chapter, you will learn how to build Random Forests, our first ensemble method and a rival of XGBoost. The applications of Random Forests are important for comprehending the difference between bagging and boosting, generating machine learning models comparable to XGBoost, and learning about the limitations of Random Forests that facilitated the development of XGBoost in the first place.