Corporate Financial Distress Based On Logistic Regression & Neural Network Research

With China's rapid economic development, enterprise bankruptcy studies have been not only more and more attention, but also the focus of attention has focused on China's listed companies. Listed companies are the foundation for the development of the securities market, their behavior and performance directly determines the rise and fall of the stock market. The study on listed companies' financial distress is being more and more attention. Along with the correlation theory development and the research technique improvement, this research unceasingly obtains the promotion, but other benefit sides impelled academic circles research step. Goes on the market the corporate finance failure forecast, it not only has a higher academic value, moreover has the huge society application value. At present the gradually market China economy, urgently is needed to consummate the economical forecast method, establishes the economical forecast system. For enterprise managers, investors, creditors and other stakeholders who evaluation of the operational situation, investment value of the credit situation, the study on financial distress is great practical significance.Based on the domestic and foreign scholars' research, analyze the principle and modeling thoughts of Single models and Hybrid Prediction model, and select 20 financial indicators to build prediction model. Because previous studies using the "proportion" to choice the samples, that is subjective. Therefore extracts three groups of samples in 2007 China's 1413 listed companies, which are used in training and testing. By comparing the analysis of the Single models and Hybrid model, the results showed that the combination forecast models' accuracy and robustness is the best.