Bagging, Boosting, and C4.5

Breiman's bagging and Freund and Schapire's boosting are recent methods for improving the predictive power of classiier learning systems. Both form a set of classiiers that are combined by voting, bagging by generating replicated boot-strap samples of the data, and boosting by adjusting the weights of training instances. This paper reports results of applying both techniques to a system that learns decision trees and testing on a representative collection of datasets. While both approaches substantially improve predictive accuracy, boosting shows the greater beneet. On the other hand, boosting also produces severe degradation on some datasets. A small change to the way that boosting combines the votes of learned classiiers reduces this downside and also leads to slightly better results on most of the datasets considered.