Lecture 4 (Part 1) - Decision Tree and Random Forest and Regression Tree - Machine Learning Course
Machine Learning course taught by Dr. Mohamed-Rafik Bouguelia, at Halmstad University, Sweden.
This lecture presents the Decision Tree Classifier, the Regression Tree, and Random Forest. First, it defines what a decision tree is and how to use it to make predictions. Then, it explains how to construct (train) a decision tree by minimizing the impurity (or maximizing the information gain), and presents the Entropy measure and the Gini impurity. Second, it presents the Regression Tree and how to construct (train) it by minimizing the mean squared errors. Then, shows the advantages and disadvantages (pros and cons) of decision trees (and regression trees) and how to avoid overfitting. Finally, it presents an ensemble method (Random Forest) that combines several diverse decision trees, using Bagging.