An Evaluation of C4.5 and Fuzzy C4.5 with Effect of Pruning Methods
Tayyeba Naseer, Sohail Asghar · IGI Global eBooks · 2017
Classification is a supervised learning technique in data mining classify historical data. The decision tree is easy method for inductive inference. The decision tree induction process has three major steps – first complete decision tree is constructed to classify all examples in the training data, the second is pruning this tree to decrease misclassification rate and the third is processing the pruned tree to improve the classification. In this chapter, the empirical comparison of pruning the tree created by C4.5 and the fuzzy C4.5 algorithm. C4.5 and Fuzzy C4.5 decision tree algorithms are implemented using the JAVA language in Eclipse tool. In this chapter, first decision tree is built using C4.5 and Fuzzy C4.5 and five famous pruning techniques is used to evaluate trees and the comparison is achieved between pruning methods for refining the size and accuracy of a decision tree. Cost-complexity pruning produce the smaller tree with minimum increase in error for C4.5 and Fuzzy C4.5 decision trees. Request access from your librarian to read this chapter's full text.