Font Size: a A A

Personal Credit Card Application Risk Assessment Model

Posted on:2009-11-13Degree:MasterType:Thesis
Country:ChinaCandidate:Y Q LiFull Text:PDF
GTID:2199360245483010Subject:Quantitative Economics
Abstract/Summary:PDF Full Text Request
Along with the fast fierce development of our country financial market, the credit card business of our country now is placed in booming development stage .Currently credit risk control becomes the finance risk control of an importance contents. The issue of credit card essence to belong to a letter loan business. Passing the analysis credit card business risk to the becoming because of, establishment the perfect risk control mechanism, and dig out efficiently of risk control technique, for dissolving the business risk of the commercial bank credit card to have very importance of reality meanings. The current problem our country commercial bank faces is credit risk, not only influence commercial bank of steady management, but also stirs up many inside concealed of finance risk. The credit risk endangers the financial market biggest, for it direct influence modern economy life medium of various activity, and influence arrives macro view decision and society economy of development of the nation. Therefore, establishment an efficiently of estimate model, accurately analysis and confidence commercial bank management risk are the effect guards against credit risk for exaltation efficiency that believe loan risk's supervision, for assurance our country finance the system be steady, efficiently circulate, realization economy of can keep on develop have importance meanings. Compared with the abroad bank credit card business, the risk management level of credit card business of our country commercial bank is lower, the management means and methods comparison are fall behind. Lacking a set of valid of application risk valuation method is one of the main factors, which bar personal credit card business further open an exhibition.This text main is around how to control application credit card of credit risk passing the calculation of model. The form need to fill when you application credit card, which altogether can induce for 17 factor. Passing pair of this 17 factors value importation model, then get 1 to output value in the end. According to this output a value, we can judgment the size of credit risk, and decide whether hair card or not, and the limiting is how much etc.Currently the controlling application risk model main have layer analysis method model, according to the model of the method process of decision tree, and VaR method etc. However, the above-mentioned 17 times with some factors are not unavoidably whole, so the usage of these models more easily comes amiss. So we choose the nerve network model, which can permit wrong ability ratio stronger. Past origin the selected works chose to permit wrong ability ratio stronger nerve network model.The point of research of tradition of nerve network model is around how to assurance the importation of network, and output a layer Wei number of set up the mold problem. However, when we research how to set up a mold of complications the system, Because of the excessive impact factor, we can't assure the redundancy factor with useful factor, and can't simple the factor of the importation simplification. So when the importation information space Wei number bigger, the network not only structure complications, and the training time also very long, lower network function thus, influence calculation accurate degree.Therefore, this text's trying exploitation layer analysis method is a BP nerve network of front processing, passing to have already have of expert judgment, comparison, evaluation etc, and means many changes quantity of importance the degree amount turn, which result be a BP nerve the importation of the network value with let up BP nerve a network of structure of complexity, shorten training time thus, and full make use of BP nerve the network strongly permit wrong ability and anti-interference ability, exaltation the efficiency of the model.The combination passed to make use of layer analysis method(AHP) and BP nerve network to combine together, evaluation the method carry on reputation grade valuation to the credit card applicant, we can get 1 of output value ,which lies between 0 to 1. Basis this value can elucidation the applicant's credit risk condition. This can in aid of bank card center measure whether rightness should customer hair card, and the limiting is how much etc. If we do like this ,we can availably lower the credit risk of credit card thus.
Keywords/Search Tags:credit card, credit risk, layer analysis method, the AHP-BP nerve network
PDF Full Text Request
Related items