Comparative Study of C4.5 and Boosting using Decision Tree Learning Algorithm

Abstract

Data Mining aims to discover novel, interesting, and useful knowledge and patterns from databases. Classification is a data mining technique which addresses the problem of constructing a predication model for a class attribute given the values of other attributes and some examples of records with known class. Decision tree are one of the most well-established classification methods. They are so popular because their ability to handle nisy data, their comprehensibility, and their capability to learn disjunctive expression. One of the most popular decision tree construction algorithm is C4.5. The idea of esemble methodology is to build a predictive model by integrating multiple models for better generalization error. It is well known that ensemble method can be used for improving the predictive performance. Boosting is one of the methods for build ensemble of classifier. This paper compare s the popular C4.5 and boosted C4.5 for their prediction accuracy using holdout method.