Comparation And Analysis Of Financial Early,Warning Between Entropy-Neural Network Model And Logistic Regression

In 2015, the first ever mandatory leverage adjustment resulted in a large evaporation of assets. The entire economy paid a heavy price. The national economic performance exposed the depressed economic situation. Faced with such a volatile economic situation, listed companies, which had been presented with the special treatment or special transfer, have reached as many as 110 cases.As a representative of National Productivity, manufacturing listed companies not only need to predict and regulate the development of the situation, but also need to prevent themselves from falling into financial distress now. Thus, the financial crisis anticipation would be more meaningful.Based on the previous research, this paper applied information entropy to screen indicators, which would be applied to the artificial neural network model later. Then this essay compared the difference and similarity between artificial neural network model and the traditional statistical model.Firstly, this paper did the study of the models and indicators from the former literature. Second, it expounded the relevant theories. Third, it designed the model indexes, and put forward the hypotheses, and determined the sample range.Finally, we use the factor analysis to establish logistic model. And then, we established the entropy-ANN model, which would be compared with the logistic regression model. The adjunctions of new indicators were the biggest innovation of this article. And the empirical results demonstrated the superiority of the new entroy-neural networks.