Reserch On Financial Crisis Early Warning Based On BP Neural Network Model

Today’s international competition is science technology and talent’s competition. As high-tech enterprises and science technology personnel carrier to become an emerging force and an important part of national economic development.In 2014,China’s innovation activities continued to be active,high-tech enterprises is continuing with rapid development. However, due to the special nature of its business type,We will face many challenges and risks in the development process, such as R & D is not enough, the market share decline and higher research investment risk; and Shanghai and Shenzhen from 2012- 2014 the high-tech enterprises listed company data, the number of ST companies continued to rise. This year,We will make efforts to promote the annual plan of "Thirteen Five" scientific and technological innovation, It is not only an opportunity but aiso a challenge for the development of high-tech enterprises.In order to achieve the strategic high-tech development, relying on the strength and resources of change, rational and effective conduct the risk aversion of the high-tech enterprises and prevention of the outbreak of the financial crisis,it has become the current research focus of scholars and business management authorities concerns.High-tech enterprises should establish appropriate financial crisis early warning system and make in advance to deal with according to the danger signals emitted.Firstly, the financial crisis and the financial crisis early warning theory expounded the theory, combined with the current status of the development of high-tech enterprises, systems analysis of the need for high-tech enterprises will be the financial crisis early warning; then select the indicators were based on certain principles from solvency, seven aspects of its operation capacity selected 28 financial early warning indicators. To make warning indicators for model construction, this paper, statistical analysis software to optimize early warning indicators.On the basis of theoretical analysis, the paper BP neural network to build high-tech enterprises of the financial crisis early warning model, predicted by the training samples and testing samples of T-1 and T-2 year’s data to predict the T year’s financial condition, getting the final statistical conclusions: T-1 year’s prediction accuracy is 88.5%, T-2 year’s prediction accuracy is 83.45%, the result is desirable. Through theoretical analysis and modeling, this paper obtained the following conclusions:(1) The financial crisis early warning index system constructed to predict the effect can play better, more fully reflect the financial situation of enterprises.(2) Application of BP artificial neural network model to predict the financial crisis prediction accuracy is high, there is a strong promotion of practical value.(3) This article tech company crisis early warning model constructed in between the high-tech enterprises to promote, provide a reference value for the enterprise management authorities.