Empirical Study on the Effectiveness of Bagging and Boosting

摘要：

Abstract: This paper deals with issue of the effectiveness of bagging and boosting. How effective is bagging and boosting in improving classifier’s accuracy? This paper try to answer this question by performing experiments on bagging and boosting three types of classifier, which are J48, Na.ve Bayes and IBK (one type of nearest neighbor classifier). I compare the performance increase in terms of error rate. Note that J48 are unstable learner, while both NB and IBK are stable learners. The experiment result also empirically verified that bagging and boosting works well only on unstable learner.

关键词： Bagging Boosting Classification Algorithm

作者: Yin-zhi Lei

作者单位: School of Information Resource Management, Renmin University, Beijing，P.R.China, 100872

会议类型: 国际会议

会议名称: 2011 International Conference on Information System and Computational Intelligence(2011 IEEE信息系统与计算智能国际会议 ICISCI 2011)

会议地点: 哈尔滨

会议语种:英文

页码: 300-303

在线出版日期: 2011-01-18（万方平台首次上网日期，不代表论文的发表时间）