Research On Eyaluation Method Of Enterprise Perforniance Based On Fuzzy Neural Network

As an important modern management method and means, Enterprise performanceevaluation is playing an increasingly important role in the background of economicglobalization. It not only leads the development of enterprise, but also affects theirbusiness stakeholder to make decisions.The fuzzy neural network was applied into the field of the enterprise performance;meanwhile, the financial indicators and non-financial indicators were integrated toevaluating the enterprise performance. The fuzzy theory and neural network wereapplied to minimize the influence of human subjectivity to the evaluation indexes, andfinally, the fuzzy neural network for enterprise performance was constructed, whichadvances the intelligence of enterprise performance evaluation.The main content of the paper is as follows:Firstly, the development of the methods of enterprise performance evaluation invarious stages of domestic and oversea was summarized in detail. The relevant theory offuzzy neural network, the fusion characteristics and combination mode of fuzzy logicsystem and neural network was studied systematically,.Then, the construction ideas and principles of enterprise performance evaluationindex system was put forward on the basis of previous studies, and finally the indexsystem including financial and non-financial indicators was established, whichinvolving the profit ability, assets operation ability, solvency, development ability,enterprise managers, employees, customers and business contributions to society etc. Inaddition, the dimensionless method of the evaluation indicators was studied detailedwith the aim to make the evaluation data in the same magnitude.Thirdly, in order to solve the dimension disaster of fuzzy rules, and simplify thenetwork structure of evaluation model and reduce the training time of network, weconstruct the combination model of enterprise performance evaluation based on fuzzyneural network. The BP neural network was introduced to build the financial indicatorsmodel and Takagi-Sugeno fuzzy neural network was applied to build the non-financialindicators model, then the combined model for enterprise performance evaluating wasconstructed applied the above results as input. The results of experiments show that theapplication of combination model of enterprise performance evaluation is feasible andeffective, and has good accuracy. The combination model can be used as a viable alternative solution for enterprise performance evaluation.Finally, I establish a general framework of enterprise performance evaluationprototype system based on fuzzy neural network, and introduce the function of eachmodule.