改进BP神经网络下的养老保险介入风险评估系统

Endowment insurance subsidy intervention risk assessment system based on improved BP neural network

为了准确、高效地评估养老保险补贴介入的风险产生概率与危害程度,改进BP神经网络下的养老保险介入风险的评估系统。以创建一个通用的养老保险补贴介入风险的评估环境为前提,设计包含五大模块的养老保险补贴介入风险评估系统总体框架,采用贝叶斯正则化算法优化BP神经网络的权值和学习率,并将养老保险补贴介入风险指标作为优化后的BP神经网络的输入向量,根据网络获得参数最优解,集成养老保险介入风险评估信息,实现养老保险补贴介入风险评估。实验结果表明,该系统具备较高的养老保险补贴介入风险评估效率,可有效评估出不合理的风险因素,准确评估出养老保险补贴介入的风险产生概率与危害程度。

An endowment insurance subsidy intervention risk assessment system based on BP neural network is improved to accurately and efficiently evaluate the risk probability and harm degree of endowment insurance subsidy intervention.On the basis of the premise of creating a general assessment environment for the endowment insurance subsidy intervention risk,the overall framework of endowment insurance subsidy intervention risk system including five modules is designed.The weight and learning rate of the BP neural network are optimized by means of the Bayesian regularization algorithm,and the endowment insurance subsidy intervention risk index is used as the input vector of the optimized BP neural network.The optimal solution of parameters is obtained according to the network,and the risk assessment information of endowment insurance subsidy intervention is integrated,so as to realize the risk assessment of endowment insurance subsidy intervention.The experimental results show that the system has high efficiency in risk assessment of endowment insurance subsidy intervention,can effectively assess the unreasonable risk factors,and accurately assess the risk probability and harm degree of endowment insurance subsidy intervention.