Study On Financial Crisis-predicting System Of Listed Company Of Fast Moving Consumer Goods Based On Feed-forward Backprop Neural Network

Because our country to the growing demand for fast moving consumer goods, and because the industry into the barriers is low,fast moving consumer goods production enterprise increased continuously increasing homogeneity product, market competition more hasten is intensified. Enterprise in its management process,and face a market uncertainty and risk also constantly enlarges, financial risk pre-warning has become a modern enterprise financial management the important content. Constructing accurate, effective financial risk early warning model, the timely discovery enterprise operation activities of the existing problems, and guide the enterprise take measures to avoid risk, is not only of theoretical value, but also has the important practical significance.This paper with the financial risk theory as guidance, from the domestic and foreign scholars in the basis of research achievements, the first analysis of the causes of the financial risk, expounds the financial risk pre-warning system structure, function, review the current domestic and foreign financial risk pre-warning system research results. Second, fast moving consumer goods industry is defined, expounds the fast consumption goods industry characteristics and present situation, and then the current financial risk pre-warning methods of analysis, this paper summarizes the of all kinds of the shortcomings of the methods and defects, and artificial neural network to introducing financial risk pre-warning research into the theoretical analysis. Again, this paper, from the debt paying ability, profitability, operation ability, growth ability and cash flow ability five were18indicators financial risk assessment index system, build an FMCG industry characteristics of financial risk pre-warning index system, the BP neural network model and simulation combining, attempt to establish a financial system without artificial processing, it is easy to operate financial risk early warning model. This paper use of Matlab software engineering, will be2009to2011in70to home fast consumption goods listed companies’data as sample, and according to the neural network algorithm iteration and processing, make the object-oriented modeling of the neural network and the training sample can be realized, finally network warning ability test. Experiment research training among set hidden layer neural circuits to predict the number of the chesses, and through repeated the experiment oerall predictie accuracy of90%for the model. Thus, the accurate forecast of the listed company of consumption goods fast financial risk, and for enterprise to find out the possible existence of financial risk timely take corresponding measures, the maximum avoid financial crises.