International Journal of Advanced Research in Computer Science and Software Engineering (IJARCSSE)
Classification of data plays a crucial role in the field of data mining. The size of the data decreases the performance and efficiency of classifier. The decreasing performance of classifier compromised with un-voted data of classifier. So the merging of two or more classifier is done for better prediction, such techniques are called ensemble classifier. One such approach is using bagging and boosting. This paper deals with enhancing the accuracy of prediction by presenting an ensemble method which is modification to boosting, which uses the dataset as the test set instead of the training set for error calculation during derivation of the model.