Methodological Study On Enterprise Risk Dynamic Management Based Upon Indeterminacy Theory

Enterprise risk management is to recognize, weigh and control enterprise risks in a systematic and standardized manner so as to minimize all possible losses caused by risks with minimum costs. Generally, in the actual management, there exist many uncertainties not only of risks but of subjective or improvisational factors in making decision. In this case, indeterminacy theory is usually employed to quantify and manage these uncertainties. Accordingly, on the basis of indeterminacy theory, and in line with system theory, control theory, artificial intelligence, this article aims at putting forward some dynamic methods, which may help manage enterprise risks technologically.There are four chapters in this article. In the first chapter, the significance and the outline of the study are introduced. In the second chapter, problems of enterprise risks prediction are presented and the prediction methods of the trait indices are examined and researched. And in the third chapter, the dynamic recognition and the evaluation methods are analyzed. Finally, in the fourth and last chapter, the dynamic control and methods of making decisions in enterprise risk management are studied.What are innovative in this article are as follows:1. Grey neural network is proposed to predict the trait indices of enterprise risks. The data limits of GNNM (1,1) predictive model are theoretically presented and the results here are testified to be valid. Therefore, GNNM (3, 1) predictive model is set up grounded on GNNM (1,1) and proved valid.2. Considering the problems in recognizing the levels of enterprise risks, fuzzy pattern recognition and neutral network recognition are combined, with their advantages complementary. The combination of these two recognition methods is proved valid through empirical analysis.3. In view of the evaluation indices and requirements of enterprise risks, and based on fuzzy optimum theory and system theory, a multi-layer and multi-objective fuzzy optimum dynamic composition evaluation is proposed. By this method, the system iterative equation is established and evaluation sub-systems are organically combined, making a dynamic composition evaluation and comparative analysis of the objections. In determining weight, due to their respective advantages subjective and objective evaluation are blended to make the coefficient determination of multi-index composition evaluation more reasonable.4. For the purpose of transforming and solving inconsistent problems, the system quantification method of enterprise risk control is raised in this article, on the ground of human controlling behaviors and such theories as fuzzy logic, program control theory, system programming theory, cluster analysis technology and incidence analysis technology. First, the state and space of enterprise risks are described in form of "power" and on the basis of fuzzy extension theory. Second, to avoid false recognition caused by accidental and unusual changes, the article suggests the moving deviation coefficient, achieving the goal of timely and dynamic recognition. Third, the influence coefficient is set up to measure the influence of risks control strategies over trait indices. In addition, the multi-dimension trait set and its function are introduced to show the influence, considering quantified and qualified information. Finally, with extension theory, incidence function is used to measure and evaluate the degree and effect of control strategies to select the best one.5. In view of complicated decision-making problems in enterprises, a multi-dimension and multi-objective fuzzy optimum dynamic programming model is established. Furthermore, RAGA and dynamic programming are combined for solution, which may guarantee the optimum variable.