摘要
以我国中小企业板上市公司退市风险预警为例,利用弹性反向传播算法(resilient back propagation,Rprop)和因子分析法相结合,建立了一种基于因子分析的Rprop神经网络模型。首先利用因子分析法构建包含财务变量和非财务变量的预警体系;其次运用Rprop神经网络模型对我国160家中小企业板上市公司进行退市风险预警实证分析;最后对该模型的有效性进行了实证分析,结果表明,该模型对上市公司退市风险预警的准确性比标准的BP神经网络模型和支持向量机模型分别提高了2.91%和6.09%。因此,该模型可为投资者决策提供较好的参考依据。
Taking the early warning of the de-listing risk of small and medium-sized enterpriseboard listed companies in China as an example,a Rprop neural network model based on factor analysis is established by combining elastic reverse propagation algorithm(resilient back propagation,Rprop)and factor analysis.Firstly,the early warning system containing financial and non-financial variables is constructed by using factor analysis method,secondly,the Rprop neural network model is used to analyze the risk of de-listing risk of 160 small and medium-sized enterprises in China,and finally,the effectiveness of the model is empirically analyzed.The results show that the accuracy of the model's early warning of de-listing risk of listed companies is 2.91%and 6.09%higher than that of the standard BP neural network model and the support vector machine model.Therefore,the model can provide a good reference for investors'decision-making.
作者
虞文美
方扶星
YU Wen-mei;FANG Fu-xing(College of Finance,Anhui University of Finance and Economics,Anhui Bengbu 233030,China)
出处
《齐齐哈尔大学学报（自然科学版）》
2020年第4期69-73,共5页
Journal of Qiqihar University(Natural Science Edition)
基金
安徽省教育厅高校优秀青年人才支持计划重点项目:“一带一路”对人民币国际化的总体估计、传导机制和经济互动效应研究(gxyqZD2017047)
安徽财经大学校级课题:中国金融周期的度量、成因分析和宏观经济效应研究(ACKY1728)。
