Research Of Credit Rating Model Based On Neural Network

In this paper, first, the present development and application area of Data Mining and Neural Network technology are introduced. Then discusses the present condition of credit rating system for enterprise and individual in the world, including the main method of rating and their main defects in application. Taking advantage of the learning, nonlinear processing and error correcting abilities of neural network, a credit rating model based neural network is built according to the credit index of the enterprise and individual. It amends the defects of other methods. The model construction includes rating methods analyzing, index system and rating factors setting, sample data preprocessing and structure designing of neural network based on BP algorithm. At last, an improvement algorithm of BP is given according to an actual application. The rating model supplies the foundation of the scientific management and correct decision for credit assessment organization and finance institutions. It is also a good way to solve the similar problem and use neural network in other fields.