Decision Tree Algorithm in Machine Learning
Yifan Lu, Tianle Ye, Jiali Zheng
Abstract
Machine learning has been a hot topic in artificial intelligence for quite a few good reasons. In the future, the world’s information would be too massive for us to process. Therefore, it will be necessary of machines to help us process data, do calculations, and help us make decisions. Supervised learning and unsupervised learning are the two most widely used learnings. Decision Tree is one of the most important ones in supervised learning. It is commonly used for classifying categorical as well as numerical variables. In our paper, we will first introduce the metrics used by the algorithms in decision trees. We will then discuss their application in the algorithms CART and ID3. We will also be discussing C4.5, an extending algorithm from ID3. In the end of the paper, we will give an introduction of pruning, which is a common measure to deal with over fitting that is seen in decision trees.