Multiple Boosting: a Combination of Boosting and Bagging

Classiier committee learning approaches have demonstrated great success in increasing the prediction accuracy of classiier learning , which is a key technique for datamining. These approaches generate several classiiers to form a committee by repeated application of a single base learning algorithm. The committee members vote to decide the nal classiication. It has been shown that Boosting and Bagging, as two representative methods of this type, can signiicantly decrease the error rate of decision tree learning. Boosting is generally more accurate than Bagging, but the former is more variable than the latter. In addition, Bagging is amenable to parallel or distributed processing , while Boosting is not. In this paper, we study a new committee learning algorithm, namely MB (Multiple Boosting). It creates multiple subcommittees by combining Boosting and Bagging. Experimental results in a representative collection of natural domains show that MB is, on average, more accurate than either Bagging or Boosting alone. It is more stable than Boosting, and is amenable to parallel or distributed processing. These characteristics make MB a good choice for parallel datamin-ing.