摘要
从政府投资工程代建管理模式运作方式出发,应用现代项目风险管理理论与方法,对政府投资工程实施代建制的风险来源和影响因素进行分析,运用前馈神经网络模型,根据BP算法调整其权值和阈值,进行系统训练并结合输入数据,最终得到风险估计系数作为风险管理决策依据,为政府投资工程的风险分析提供一种现代算法的新思路。
The risk sources and factors of the agent mode in government projects are analyzed. Based on BP neural network the risk coeffieients, as the foundation of risk management decision, can be obtained by adjusting the weight and threshold through BP algorithm, as well as training systematically with input data. It provides a new thought along with new algorithm for the risk a- nalysis of government projects.
出处
《武汉理工大学学报（信息与管理工程版）》
CAS
2007年第6期89-92,共4页
Journal of Wuhan University of Technology：Information & Management Engineering
基金
湖北省建设厅建设科研资助项目(K200566)
关键词
神经网络
政府投资工程
代建制
风险分析
neural network
government invested project
agent system
risk analysis
