基于聚类-粗糙集-神经网络的企业财务危机预警

发布时间：2018-08-18 16:10

【摘要】：财务危机预警是企业各方利益相关者关心的一个重要问题。由于样本的可得性问题,现有的财务预警相关研究一般只能把企业的财务状况分为ST和非ST两类,运用层次聚类将其分为健康、良好、中等、轻警、重警5种类型,从而对企业财务状况的描述更符合实际。在考虑预警指标所含信息的完备性和准确性的基础上,使用了粗糙集方法构建指标体系,弥补了传统指标体系构建时信息量冗余或不足的缺陷。最后,将5种分类作神经网络的输出层,粗糙集约简后指标体系作输入层,构建神经网络模型进行财务预警。通过粗糙集约简形成的输入层节点数以及聚类形成的输出层节点数,进而提升了神经网络的拟合效率,使其运用更为合理,并最终完成对财务状况转折点以及财务状况恶化阶段的预测。
[Abstract]:Early warning of financial crisis is an important issue concerned by stakeholders. Due to the availability of samples, the existing financial early warning research can only divide the financial situation of enterprises into St and non-St, and use hierarchical clustering to divide them into five types: healthy, good, moderate, light warning and heavy warning. Thus, the description of the financial situation of the enterprise is more realistic. On the basis of considering the completeness and accuracy of the information contained in the early warning index, the rough set method is used to construct the index system, which makes up for the defects of redundant or insufficient information in the construction of the traditional index system. Finally, five kinds of neural network are classified as the output layer of neural network, and the index system after rough set reduction is used as the input layer, and the neural network model is constructed to carry out financial early warning. By reducing the number of nodes in the input layer and the number of nodes in the output layer formed by the rough set reduction, the fitting efficiency of the neural network is improved and the application of the neural network is more reasonable. And the final completion of the financial status of the turning point and financial deterioration phase of the forecast.
【作者单位】：北京联合大学管理学院;北京联合大学商务学院;
【分类号】：F275

【相似文献】