基于神经网络技术的企业财务危机预警研究

Pre-warning study of corporations'financial crisis based on ANN technique

在综合研究的基础上提出了基于财务视角和非财务视角综合的企业财务危机识别系统,将企业划分为财务健康、灰色地带和财务危机三类,摈弃了以往国内相关研究仅仅以"ST"、"PT"公司直接作为财务危机研究对象的做法。运用主成分分析和神经网络技术对我国上市公司2005、2006年度的财务状态进行了实证研究,并利用预警模型和因素分析模型对天目药业(600671)进行了预警实例分析。

This paper aims to improve the financial crisis forecasting model based on the indicator of financial ratio and non-financial ratio.A crisis prediction model is established to distinguish three categories of enterprises according to their financial status instead of only using 'ST'&'PT'companies in researching financial crisis.With principal component analysis and ANN a research is done on the listed corporations of 2005/2006 and Tianmu Corporation(600671) is taken as an empirical example in crisis prediction research.