Bagging and Induction of Decision Rules
Jerzy Stefanowski · 2002
An application of the rule induction algorithm MODLEM to bagging is discussed. Bagging is a recent approach to construct multiple classifiers that combines homogeneous classifiers generated from different distributions of training examples. The basic characteristics of bagging and the MODLEM are given. This paper reports an experimental study of using bagging composite classifier and the single MODLEM based classifier on a representative collection of datasets. The results show that bagging substantially improve predictive accuracy. These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.