The Effectiveness of the Combination of Filter and Wrapper Feature Selection Methods to Improve Software Fault Prediction

Improving the software product quality before releasing by periodic tests is one of the most expensive activities in software projects. Due to limited resources to test modules in software projects, it is important to identify fault-prone modules and use the test sources for fault prediction in these modules. Software fault predictors based on machine learning algorithms, are effective tools for identifying fault-prone modules. Extensive studies are being done in this field to find the connection between features of software modules, and fault-prone have been done. Some of these features in predictive algorithms are like that not only did not improve the accuracy of the learning process, but also will be reduced the accuracy. In this study, due to the excellent performance of Forward Feature Selection (FS) method for effective selection of features, the initial subset of this method has been selected by using of combination of high ranking features in different Filter methods. The proposed method causes increment the speed of the convergence of feature selection as well as the accuracy improvement. The obtained results on NASA dataset with AUC criteria, indicates the effectiveness of this method in the improvement of the accuracy and the speed of software fault prediction.