Enhancing Ensemble Performance through Feature Selection and Hybridization

Ensemble has been proved a successful approach for enhancing the performance of a single classifier. But there are two key factors directly influencing the outcomes of an ensemble: accuracy of each single member and diversity between the members. There have been many approaches used in the literature to create the mentioned diversity. In this paper, we add to them a novel approach, in which classifier type variance is utilized along with feature subset diversification to create a high diversity ensemble of different classifiers and the ensemble is optimized using a multi-objective evolutionary algorithm. The suggested approach outperformed existing ones in experiments conducted on some standard datasets.