Research On Evaluation And Early Warning System Of Industry Security

ABSTRACT:Evaluation and early warning system of Industry Security is based on scientific and comprehensive analysis. According to the indexes variation, it should warn timely and accurately in order to make reasonable suggestions. Accurate industry security early warnings can prevent hidden trouble, resist economic risk and maintain industry security. The establish of evaluation and warning system of industry security is to evaluate the various parameters, analyze the industry's safety, then focus on the key and sensitive parameters, assess the impact on the industry, publish the information of substantial damage and threat, so as to safeguard industry security, adjust the industry structure, formulate various measures, and realize industry protection work in advance.In this paper, industry is seen as a system and is analyzed with theoretical analysis and mathematical statistical methods. Firstly, it describes the influence factors of industrial security, classifies the risk indicators and builds the industry security evaluation system. Secondly, it calculates the weight of every single index in detail with principal component analysis method, and builds industry security evaluation model. Lastly, it analyzes critical value of the indexes, builds industry security early warning model, and predicts the tendency of industry security with BP neural network.It combines principal component analysis and artificial neural network based on the combination of theoretical analysis and practical application to evaluate and monitor industry security and make early warning, to find the changes of different levels clearly, to have insight of the changes in the development and operation of industry. These will be useful for government to take measures to withstand risks timely. Finally, the author does empirical research on financial industry, analyzes this industry comprehensively, predicts its tendency, and validates the feasibility and effectiveness of evaluation and early warning model.This article totally includes 5 Figures,26 tables and 51 references.