The project will give knowledge with the implementation of decision tree. Decision Tree for Classification

The project will give the knowledge of implementation of decision tree (a popular classifier) and the procedure of evaluating the performance a classifier. In this project, the decision tree is a binary tree, i.e., each node has at most two children. For simplicity, you can assume that all attributes in a given dataset are binary. Specifically, the term project contains two parts. In the first part, you will first need to construct a randomized decision tree. The construction procedure is a simplified implementation of the tree growth algorithm. Then you will evaluate the classification performance of the randomized decision tree by 2-fold cross validation. In the second part, you will construct a gain based decision tree by modifying two subroutines in the first part. As well as the first project, you can useWeka library in your Java implementation. Especially, you can apply [login to view URL] package for the necessary work on data preprocessing. Weka provides implementations of several different versions of decision trees. These implementations can be used as references to help you finish the first part of the project. But it is important to note that our implementation is different from the implementations of Weka decision trees