基于神经网络方法的财务风险识别

Financial Risk Detection Based on Neural Network Method

利用上市公司披露的信息数据库为平台,将神经网络方法应用于财务风险识别。实证研究结论表明:不仅可以把模型的仿真度提高到100%,而且显著提高了财务状况特征识别的准确率,克服了国内以往的数据挖掘研究重理论轻实践的缺点,展现了基于神经网络方法的知识发现过程。

In this paper we apply neural network prediction method to financial risk detection on the basic of the database of listed companies＇ released information. The empirical study indicates that the company financial situation detection model based on neural network not only bring the fidelity up to 100 %, but also significantly improve the accuracy ration of detection. Not like the other data mining researches made by Chinese scholars, our work pays more attention to practice application than to theory analysis. A whole procedure of knowledge discovery of database through neural network method is demonstrated in the paper.