Font Size: a A A

A Rival Style NN Applied Research In Enterprise Seed-credit Sort

Posted on:2008-11-23Degree:MasterType:Thesis
Country:ChinaCandidate:Z YanFull Text:PDF
GTID:2189360242472571Subject:Computer technology
Abstract/Summary:PDF Full Text Request
Generally, the credit is one kind of important intangible asset. No matter to the individuals or the enterprise, good credit is the manifest of qualifications and property. Under socialist market economy condition, consummate social credit system is the basic system of modern market economy. "CCP's certain decisions about consummating socialist market economy system" in the Third Plenum of the 16th CPC Central Committee said: The Social credit system bases on moral, property right and law is not only the requirement to build modern market system but also the necessary strategy to standardize the order of market economy.From the classification of enterprise credit, especially thinking about the bank's inspection to the credit, the article analyzes and judges the following service. Modern bank risk mainly includes: credit risk, market risk, operation risk, fluid risk, country risk, prestige risk, legal risk, strategic risk. Among them, credit risk is the most important. The article makes enterprise credit model according to domestic and foreign correlative data and gets the result with the help of SOM and LVQ. At the same time, it also uses Fuzzy theory to make the conclusion understood easily and validate the model and arithmetic in practice.Because SOM arithmetic is a deputy for study without mentor, it can make immense effect in the early days when the system is founded and run with incomplete accumulated experience. However, along with the experience and the data accumulated, the experience data can't be utilized sufficiently. For taking full advantage of the experience data, fuzzy diagnosis in LVQ arithmetic can solve it. So after combining SOM and LVQ, the whole arithmetic is much closer to the fact, the data result is nicer. Then it can win better result.
Keywords/Search Tags:Enterprise Credit, Credit risk, Self-organize character mapping network, Learning vector quantity NN
PDF Full Text Request
Related items