Heterogenous Defect Prediction using Ensemble Learning Technique

Abstract

Software Defect Prediction (SDP) is the most practically used approach in the testing phase of the software development life cycle (SDLC) which helps to find out the defected module prior to testing or releasing the product. This study intends to predict the defects through an improved Heterogeneous Defect Prediction approach based on Ensemble Learning Technique which consists of 11 different classifiers. This study analyses the way Ensemble Learning technique which includes the combination of both supervised and unsupervised machine learning algorithm helps in predicting the defect proneness of modules. This technique has been applied on historical metrics dataset of various projects of NASA, AEEEM, and ReLink. The data set is sourced from the PROMISE repository. The performance of the obtained models is critically assessed using the Area under the Curve, precision, recall, f-measure. Experiment results shows that our method is comparable to the existing method for defect prediction.