AN ENSEMBLE BASED CLASSIFIER APPROACH IN DATA MINING

T Ratna, A Reddy, Asst Professor
International Journal of Emerging Engineering Science and Technology   unpublished
An ensemble method needs to have base models that perform reasonably well but are nevertheless different from one another. The ensembles are constructed by the data sampling method called bagging. The pruned ensemble set is always as accurate on average over the dataset tested as the non-pruned version, which provides benefits in terms of its application efficiency and reduced complexity of the ensemble. In this paper an ensemble approach and its prune are discussed with the existing methods
more » ... existing methods like SVM, K-NN .Experimental results were evaluated.
fatcat:c5gdhfzdbjgv3d5vvbepb6f2hi