An evolutionary programming based asymmetric weighted least squares support vector machine ensemble learning methodology for software repository mining

In this paper, a novel evolutionary programming (EP) based asymmetric weighted least squares support vector machine (LSSVM) ensemble learning methodology is proposed for software repository mining. In this methodology, an asymmetric weighted LSSVM model is first proposed. Then the process of building the EP-based asymmetric weighted LSSVM ensemble learning methodology is described in detail. Two publicly available software defect datasets are finally used for illustration and verification of the effectiveness of the proposed EP-based asymmetric weighted LSSVM ensemble learning methodology. Experimental results reveal that the proposed EP-based asymmetric weighted LSSVM ensemble learning methodology can produce promising classification accuracy in software repository mining, relative to other classification methods listed in this study.