Prediction of Software Bugs Using FCM Fused Artificial Neural Network Classifier

Abstract

The software engineering technique makes it possible to create high-quality software. One of the most significant qualities of good software is that it is bug-free. Bug detection and prevention aims to provide high-quality software that reduces the cost and time required to fix a defect, increases productivity, and allows for complete customer satisfaction. While it will be impossible to eradicate all flaws, it will be able to reduce the number of problems and their negative influence on software quality. This may be accomplished by putting in place a bug prediction mechanism that focuses on increasing software quality by lowering defect density. To extract usable information, soft computing approaches such as data mining and machine learning algorithms can be used. Soft computing approaches have been used to solve the problem of software bug prediction in a number of researches. The ability to predict necessary bugs considerably sooner in the development process is a major incentive for this study. The purpose is to stimulate people's interest in software quality testing and encourage them to employ fuzzy c means clustering (FCM), artificial neural network (ANN), and machine learning approaches to solve challenges in the field of software project success prediction. The FCM fused ANN used for bug prediction model in this study to predict software reliability. With an accuracy of 98.237percent, this research effort exhibits the usefulness and applicability of FCM fused ANN technology.