Enhanced Algorithm Performance for Classification Based on Hyper Surface using Bagging and Adaboost

To improve the generality ability of Hyper Surface Classification (HSC) , Bagging and AdaBoost ensemble learning methods are proposed in this paper. HSC is a covering learning algorithm, in which a model of hyper surface is obtained by adaptively dividing the sample space and then the hyper surface is directly used to classify large database based on Jordan Curve Theorem in Topology. Experiments results confirm that Bagging and AdaBoost can improve the generality ability of Hyper Surface Classification (HSC) in general. However, its behavior is subject to the characteristics of Minimal Consistent Subset for a disjoint Cover set (MCSC). Usually the accuracy of Bagging and AdaBoost can not exceed the accuracy predicted by MCSC. So MCSC is the backstage manipulator of generalization ability.