Application of neural network for predicting software development faults using object-oriented design metrics

Abstract

In this paper, we present the application of neural network for predicting software development faults including object-oriented faults. Object-oriented metrics can be used in quality estimation. In practice, quality estimation means either estimating reliability or maintainability. In the context of object-oriented metrics work, reliability is typically measured as the number of defects. Object-oriented design metrics are used as the independent variables and the number of faults is used as dependent variable in our study. Software metrics used include those concerning inheritance measures, complexity measures, coupling measures and object memory allocation measures. We also test the goodness of fit of neural network model by comparing the prediction result for software faults with multiple regression model. Our study is conducted on three industrial real-time systems that contain a number of natural faults that has been reported for three years (Mei-Huei Tang et al., 1999).