Analysis of Attribute Selection Method based on IDOE in Software Fault Prediction

摘要：

　　This paper presents a new attribute selection method based on Importance De-scending Order Exclusion(IDOE)to predict software fault.By using public datasets CM1 and JM1 from PROMISE repository,six machine learning algorithms-Na(i)ve Bayes,LibSVM,Logistic,Multilayer Perceptron,SMO and Random forests integrated in Weka are used for analysis.Firstly,the GeneticSearch and Ranker search method is chosen to select attributes.Secondly,these machine learning algorithms are used in order to predict precision,recall and F-measure of software modules.Then the IDOE method is used to update the attribute datasets.Finally,the different iteration experiment results are com-pared and analysed.

关键词： IDOE attribute selection software fault prediction

作者: Xin Zhao Wenyi Zhang Ning Ai

作者单位: CSSC System Engineering and Research Institute,Beijing 100094,China

会议类型: 国际会议

会议名称: 2018 3rd International Conference on Computer Science and Information Engineering (ICCSIE 2018) 2018第三届计算机科学与信息工程国际会议(ICCSIE 2018)

会议地点: 西安

会议语种:英文

页码: 557-566

在线出版日期: 2018-09-21（万方平台首次上网日期，不代表论文的发表时间）