Application of RBF Neural Network in Risk Assessment of Software Project
Jun Liu · Jisuanji fangzhen · 2011
Research on software project risk assessment problem.Because the software project has uncertainty factor,the traditional models cannot accurately set up software project risk assessment model,resulting in software project evaluation large error.In order to reduce software project risk evaluation error,a software project risk assessment model is proposed based on RBF neural network.The model constructs a software project risk assessment index collection which is divided into several statures,uses Delphi method to evaluate index score,and then is trained and modeled by the RBF neural network to get the best software project risk assessment model.Finally,in MATLAB platform,the model is tested by the software project risk assessment examples.The simulation results show that the model improves the accuracy of software project risk assessment and reduces software project risks.