Research And Application Of Computational Intelligent Technique For Economic Early-warning

Economic early-warning system is a new mode for modern management. And its final goal is to make sure that the enterprises is operating in the safe management. So research results will directly relate to the correct understanding and judgment of enterprise performance. However, traditional method often depends on expert experience and the simple model and can't meet the objective requirements of the actual economic early-warning.With the development of statistical theory in mathematics, people have started to use Computational intelligent technique to solve the difficult problem in recent years. Based on the index system of economic warning and index design principles, economic data is analyzed by series of Computational intelligent technique such as neural network, rough sets theory and Support Vector Machine. Through the establishment of the corresponding mathematical models to explain more complex economic data, the result is good.From the principles of economic warning, the thesis introduces theoretical base of Computational Intelligent technique, from artificial neural network to Support Vector Machine, as well as some important neural networks, rough sets and Support Vector Machine theory. Then it explores the artificial neural network warning method and the application of RBF network, BP neural network and Support Vector Machine in economic warning. At the same time, this thesis makes comparison between each kind of network structure. Finally, based on the previous presentation, the thesis describes contrastive research of computational intelligence in economic early-warning system. Meanwhile, it makes comparison between SVM and neural network from theory to application. And eventually it proves the possibility of Computational intelligent technique in the application of economic early-warning.