The Application Of Artificial Neural Networks In The Financial Sector Credit Risk Assessment

Credit is the basis of market economy. National credit, enterprisecredit and personal credit constitute the whole credit system of a nation.These components play important roles in the development of market economy.Personal credit system is the foundation of social credit system. Goodpersonal credit system is an effective guarantee of establishing the economicorders of socialistic market and taking precautions against financial risks.A complete personal credit system can greatly push the development ofconsumption credit. Enlarging internal demand and encouraging consumption areimportant measures to keep our country' s national economy growing. With thedevelopment of the credit economy, the importance of the personal credit isevidence. How to value a person' s credit, in order to develop the consumptionloan and private financial operation, is an urgent problem in the course ofthe financial area development, and also a question for discussion in ournation' s credit management system.This thesis begins in analyzing theoretical problem of credit risks withreference of overseas experiences in setting up credit risks evaluation system.Then, incorporate the our country' s current situation and peculiarity ofcredit risks evaluation, taking advantage of the learning, nonlinearprocessing and error correcting abilities of B-P neural networks, buildinga rating model of credit risks evaluation. It amends the defects of othertraditional methods. The model construction includes rating methods analyzing,index system and rating factors setting, sample data preprocessing andstructure designing of neural networks based on B-P algorithm.The rating model supplies the foundation of the scientific management andcorrect decision for credit assessment organization and finance institutions.