Enhanced Cost Sensitive Boosting Network for Software Defect Prediction

plays an important role in reducing the costs of software development and maintaining the high quality of software systems. The early prediction of defectproneness of the modules can allow software developers to allocate the limited resources on those defect-prone modules such that high quality software can be produced on time and within budget. It is a great challenge to address the class-imbalance and highdimensional data problems of software defect prediction. In this paper, three cost-sensitive boosting algorithms are analyzed to boost networks for software defect prediction. Most of the previously developed prediction models do not consider this cost issue. The cost sensitive prediction technique is considered as an effective means for the optimization of quality assurance activities. Cost will never be an independent term because there are too many variables involved in the calculation of cost estimation such as human, technical, environmental and political factors. When compared to cost-blind classifiers, the proposed costsensitive boosting methods considering cost information in both feature selection and classification stages provide better solutions to deal with the class imbalance and high-dimensionality problems in SDP. Software Defect Prediction is extremely essential in the field of software quality and software reliability. These systems can be used to achieve timely fault prediction for software components. Then the software quality assurance team can utilize the predictions to use available resources for obtaining cost effective reliability enhancements.