为了确定软件开发项目中不确定因素的影响,提出基于BP神经网络的软件项目风险评估模型。首先,构建了软件项目风险识别的TEMP(technology、environment、management、process)模型;其次,在TEMP识别模型基础上建立了包括17种风险指标在内的软件项目风险评估指标体系;再次,利用BP神经网络建立了风险评估模型;最后,通过MATLAB实例证明该风险评估模型的有效性和可行性。
This paper suggested a new risk assessment of software project model based on BP neural network. Firstly, it set up a TEMP ( technology, environment, management, process) model for risk identification of software project. Secondly, built an index system of risk assessment of software project based on seventeen kinds of risk indexes. Thirdly, set up software project risk assessment model based on BP neural network. Fanily, through training of the MATIAB neural network tools and testing of test data, the research on risk assessment provided a new model for software project risk assessment.