Research on Software Defect Prediction and Analysis Based on Machine Learning. Issue 1 (1st January 2022)

• Record Type: Journal Article
• Authors: Peng, Xuemei
• Is Part Of: Journal of physics. Volume 2173:Issue 1(2022)
• Publication Date: 2022-01-01
• Subjects: Software defect prediction -- machine learning -- ensemble learning -- combination -- Eclipse prediction data set; Physics -- Congresses; 530.5
• DOI: 10.1088/1742-6596/2173/1/012043
• Languages: English
• ISSNs: 1742-6588

Abstract

Abstract: The defects of machine learning prediction technology can be more comprehensive and automatic learning model to find the defects in software has become the main method of defect prediction, selection and study of algorithm is the key to improve the accuracy and efficiency of machine learning. Comparing different machine learning defect prediction methods reveals that the algorithms have different advantages in different evaluation indicators, the use of these advantages and combining the stacking ensemble learning method in machine learning is put forward different prediction algorithm of prediction results. As software metrics and again the prediction model of software defect prediction combined machine learning algorithm is based on the experiment with the model of Eclipse, the data sets show the effectiveness of the model.