The Application Of Data Mining In The Study Of Individual Credit Risk

With the business development of bank credit, personal credit cards increasingly sharpincrease in order to be able to effectively carry out the individual credit assessment and riskprediction, more and more business is an urgent need to use scientific and efficientmathematical mining tools as aids intoto the actual business.In this paper, the domestic banking database system can not effectively use the data forintelligent analysis for business decision-making reference to this situation, the credit riskof college students, for example, the combined use of clustering, association rules anddecision tree method, the personal credibility of the model and data mining modelscombining to create a credit risk assessment system, based on data mining for thediscovery of the potential risk of personal credit history. Firstly, through the analysis ofcredit card related data elements, the establishment of the college students sample-basedrevenue model, the consumption model, and on this basis to establish a personal creditstatic optimal risk model, and model evaluation; combination of risk factor analysis, usingthe classification decision tree method, the proposed model of bank customers using creditscoring models and customer analysis; Finally, the limitations of the classification decisiontree algorithm, an innovative way to improve K-neighbor discriminant method, andthrough implementation of this method and solving, focusing on the customer breaches.The ultimate purpose of this paper is to use a mature business model for efficient analysisof credit card data, comprehensive individual risk assessment system to promote theconstruction of a personal credit system, speed up the establishment of such businesspractices to provide theoretical guidance and technical support.