Software Defect Prediction: A Survey with Machine Learning Approach

Abstract

Software defect is inevitable part of software process, product and people. Developing zero defect software is always challenge to software industry. Software defect can be defined as it is unpredictable flaws in software which increase software product cost in terms of time, resource and budget. Software defect can be detected by developer and tester but it is time consuming which might delay on time delivery of software product to the client. To recover deadline gap software company required to fill unplanned resource in ongoing project. Software defect prediction will help to solve these problems by building machine learning prediction model for software fault. In this paper, we review work done on software defect by researchers in recent years. Many of researchers published their defect prediction work in the leading journals and premier conferences. Main objective of this study is to review different approaches used to predict software defect. Furthermore, few challenges addressed in this paper which might be future research area in finding and fixing software bugs.