Ensemble Learning Model based Software Defect Detection Using Bio-Inspired Optimized Features

Abstract

software development, defects are an inherent aspect of the process and can arise at any phase, including requirements gathering, coding, or testing. These defects, whether they emerge during initial planning or late in the development cycle, pose significant challenges and can affect the final product's functionality and performance. Implementing software defect prediction (SDP) techniques is crucial in managing these challenges effectively. SDP can play a vital role in reducing associated costs by identifying potential issues early, thus allowing for more efficient allocation of resources and targeted testing efforts. This paper has proposed a ensemble learning model to identify the defects in the software at early stage of development. In order to improve the software defect prediction accuracy feature optimization was done by using bioinspird algorithm. Experiment was done on real dataset having data of multiple software. Results shows that use of ensemble model with optimize feature has increases the detection accuracy.