Research On Firm Scale Based On Fuzzy Nerve Network To Listed Companies In China

Enterprise scale is the important organization foundation of national economy. Whether it is reasonable or not directly relates to the rationalization of industrial structure and the efficiency of storage quantity capital disposition, and then relates to the level and the quality of economy growth. This study takes the listed companies in Shanghai and Shenzhen as the research object, and collects the data from 2002 to 2004 of 974 listed companies which distributed in 19 industries. In particular, it traverses the scale of the vehicle industry and real estate industry. Based on the economies of scale theory such as production function theory, this study uses the nerve network and the fuzzy system to carry on research on the enterprise scale of Chinese listed companies. Because the artificial neural network has particular advantages, such as highly non-linear mapping ability, massively parallel handling ability, distributional memory, self-organizing, self-studying and self-adapted; also the fuzzy system can carry on discursion and decision-making of the human knowledge equally like the human brain, use the human knowledge effectively and deal with the situation which is not precision or unstable. This study uses these two methods to establish the fuzzy nerve network model of enterprise scale. And the model is used to carry on quantitative analysis of the enterprise scale of Chinese listed companies. At last, the results are compared with the results from traditional linear regression method, and then conclusion is provided.This paper is divided into five parts. In the introduction, it introduces the research actualities and problems of enterprise scale at home and abroad, the soft computational method, and the research object, origin of data, research content and the method; moreover, the soft computation theory is proposed to be used to carry on empirical research on enterprise scale of Chinese listed companies. In chapter 2, the estimation index system of enterprise scale is established. It uses factor analysis to select the index, and then carries on the explanation to it. In chapter 3, it selects fuzzy nerve network model of enterprise scale of Chinese listed companies. It introduces the production function which is used to establish the fuzzy nerve network model, then determines the structure of fuzzy nerve network and describes it. In chapter 4, it introduces the BP algorithm of the fuzzy artificial neural network. First, it introduces the BP algorithm; next, it introduces the BP algorithm's discursion process; finally, it provides the discursion process of enterprise scale fuzzy nerve network BP algorithm. In chapter 5, it carries on empirical research on the enterprise scale of the vehicle industry and real estate industry, and compares the results with which from the traditional linear regression method, then provides the conclusion. At last, it indicates the insufficiency of this study and the direction for the future study.