Software Defect Prediction And System Development Based On Machine Learning

In order to improve the reliability and reduce the development cost,software defect detection plays an important role in the process of software development.With the rapid development of computer technology,traditional detection methods have been difficult to meet the existing requirements.Therefore,more and more researchers apply data mining technology to the field of software defect detection.The software defect prediction technology based on machine learning can not only detect the defects of the existing software but also predict the potential defects by extracting the corresponding software data sets,which can improves the software quality and the efficiency of the development.For the software defect detection,the samples of defective software will be much less than defect-free software samples,which lead to the serious class-imbalance problem,in addition,there is a high degree of redundancy.How to solve this problem becomes a critical issue for software defect detection.A study of multi-random undersampling and POSS method for software defect prediction would be used to deal with them.After each sampling,the feature subset is obtained by feature selection algorithm,and then the feature set is obtained by using the mean value method.The experimental results show that: This method could effectively improve the performance of software defect prediction.In order to present a more intuitive,clear and accurate process of software defect prediction based on machine learning,we designed and develop the visualization system based on B/S model,which includes data preprocessing,feature selection,and classification,and presents them in graphical form.Besides,to improve the scalability of the system,the system provides an extensible interface for users to upload their own algorithms.In summary,the algorithm proposed in this paper can effectively solve the problem of unbalanced,and the visualization system has practical value in the field of software defect classification.