A clustering approach for software defect prediction using hybrid social mimic optimization algorithm

In this information era, software usage is intertwined with daily routine work and business. Defects in software can cause a severe economic crisis. It is a crucial task in the software industry to be able to predict software defects in advance. Software Defect Prediction (SDP) aims to identify the potential defects based on the software metrics. A software module is a software component(piece of program) that contains one or more procedure. In this study, we propose a clustering approach for grouping the software modules. This work proposes a hybrid elitist self-adaptive multi-population social mimic optimization technique (ESAMP-SMO) for clustering the software defect modules. The objective function (fitness function) of the proposed study minimizes the intra cluster distance and maximizes fault prediction rate. In this study, we used the three popular benchmark NASA datasets (CM1, JM1 and KC1) for the experimental work. The performance comparison analysis shows that the proposed clustering technique outperforms the other competitor approaches.