期刊名称：International Journal on Electrical Engineering and Informatics
印刷版ISSN：2085-6830
出版年度：2018
卷号：10
期号：3
页码：447-465
DOI：10.15676/ijeei.2018.10.3.3
出版社：School of Electrical Engineering and Informatics
摘要：Software engineering activities comprise of several activities to ensure that thequality product will be achieved at the end. Some of these activities are software testing,inspection, formal verification and software defect prediction. Many researchers have beendeveloped several models for defect prediction. These models are based on machinelearning techniques and statistical analysis techniques. The main objective of these modelsare to identify the defects before the delivery of the software to the end user. Thisprediction helps project managers to effectively utilize the resources for better qualityassurance. Sometimes, a single defect can cause the entire system failure and most of thetime they drop the quality of the software system drastically. Early identification of defectscan also help to make a better process plan which can handle the defects effectively andincrease the customer satisfaction level. But the accurate prediction of defects in software isnot an easy task because this is an indirect measure. Therefore, it is important to findsuitable and significant measures which are most relevant for finding the defects in thesoftware system. This paper presents a feature selection based model to predict the defectsin a given software module. The most relevant features are extracted from all features withthe help of seven feature selection techniques and eight classifiers are used to classify themodules. Six NASA software engineering defects prediction data sets are used in this work.Several performance parameters are also calculated for measuring the performance andvalidation of this work and the results of the experiments revealed that the proposed modelhas more capability to predict the software defects.
关键词：software fault prediction; classification techniques; feature selection; f;measure;area under curve
