基于椭圆概率神经网络的企业财务危机预警

在国际经济一体化的大环境下，随着企业竞争的日趋激烈，企业进行财务危机的预警显得尤为重要。基于椭圆概率神经网络的财务危机预警方法是概率神经网络的改进，考虑代表输入变量重要性的变量权值、代表样本有效范围的核宽倒数及代表样本可靠程度的数据权值，利用该方法对企业财务危机进行预测。实证分析结果证明：此方法的优越性有较高的预测准确性，能够为中国资本市场的发展提供很大的帮助。

Research on Pre-warning of Enterprise Financial Crisis Based on Oval Probabilistic Neural Network

As the fierce competition among enterprises is becoming more and more obvious, the early warning of the financial crisis becomes particularly important in the international economic integration environment. In this paper, the financial distress prediction method is based on elliptic probabilistic neural network. It＇s the improvement of the probabilistic neural network. It considers the variable weight that represents the importance of the input variables, the width reciprocal that represents the effective range of samples and the data weights that represents the reliability of samples. Finally, we obtain the higher prediction accuracy of this method through empirical analysis. And it can help develop China＇s capital market.