SOFTWARE DEFECT PREDICTION USING MACHINE LEARNING TECHNIQUES

Abstract

Software flaws can result in large financial losses, decreased user’ happiness, and a general decrease in the dependability of software systems. Therefore, early in the software development lifecycle defect, detection and mitigation are critical. This paper provides an overview of machine learning-based software defect prediction. Exploring the use of machine learning algorithms to forecast and categorize software problems based on historical data and other software metrics is the goal. In addition to discussing the difficulties in defect prediction, this study provides a thorough analysis of the most recent machine learning methods used to this problem. Additionally, it sheds light on the feature selection, evaluation criteria, and data preprocessing methods frequently used in software defect prediction studies. The article concludes with a comparison of various machines learning methods and how well they perform in forecasting software flaws, highlighting the advantages and disadvantages of each method. The results of this study help to advance our understanding of software defect prediction using machine learning and give researchers and practitioner’s advice on how to select the best tools for their individual needs.