Research On Company Financial Crisis Prediction Based On Rough Set And Neural Network Technology

With the global economic integration of the big step and Our market economy has developed rapidly,Increasingly fierce competition among company, Market in which an unlimited business opportunities, surprises of the risks and crisis. Uncertainty and complexity of the Market economic environment are growing, This will probably make company in the financial crisis. Company in the financial crisis, not only endangers the company itself to the survival and development and spread to the investors and the interests of all sectors of society.Therefore, building the necessary financial crisis early warning mechanism and to establish a powerful financial crisis early warning system has very important actual significance. This paper introduces a number of domestic and international financial crisis early warning research of the relevant documents.Careful comparison and analysis of the current financial crisis early warning method. At present, about the financial crisis early warning method of the study includes qualitative analysis methods and quantification analysis methods introduced after two categories. however,quantification analysis methods used most widely.quantification analysis methods mainly include univariate judge model, multiple linear judge model, logistic judge regression model and ANN.However,research are mostly confined to static method. The text proposed based on Rough set assembly with the neural network of dynamic financial crisis early warning method. The method for the financial crisis early warning research provides new ideas, and can better informed and perfecting corporate financial crisis early warning of the theory and methods.Rough set theory and neural network combined with good synergy, has aroused extensive concern, both domestic and foreign scholars. This paper hope that combine the advantages and use of nerve network the processing ability of nonlinear systems, with coarse of knowledge about jane theory as a nerve network model, before dealing with units were set theories about the nerve network input financial indicator for property about jane and simplify the nerve network of the structure.To establish an effective financial crisis early warning system.Assembly through the Rough-ANN model, establish, training and inspection, we find that the Rough-ANN model has high forecast accuracy,It has good financial crisis early warning effect and early warning ability.The establish of the Rough-ANN model has viability and effectiveness to financial crisis early warning and achieve the desired effect.