Research And Construction Of Risk Control Model Based On Credit Card Transaction Data

With the rapid development of Internet technology in the world,the application of online payment has been widely popularized.With the precipitation of massive transaction data,banks have accumulated a large amount of data,including all kinds of transaction data,integral type information,customer account information,etc.Among these transaction data,the types are diversified.Unstructured types of data bring great trouble to data analysis.The occurrence of bank transactions is inevitably accompanied by different levels of risk,so criminals through the security loopholes of banks to do some illegal transactions,such as stolen brush,fake brush,set of points,etc.,such phenomena occur frequently,which bring huge losses to the banking industry and negative impact on the social atmosphere.Therefore,it is urgent to build a set of risk control model based on big data technology and artificial intelligence mining market credit card transaction data.In view of the above requirements,this paper studies the risk control model based on credit card transaction data,and provides a safe and reliable basis for the bank credit card industry.This paper analyzes and determines the main functions of the risk control model,including: user portrait label design,user portrait data run batch,feature data dimension reduction processing,model algorithm selection,risk control model construction,the practical application of risk control model.Among them,the bank credit card transaction data contains sensitive information,desensitization treatment in pretreatment,user label according to the user's information construction,the risk control model to predict whether a user for illegal trading,in this paper,the experiment as a prediction model using the algorithm of SVM and logistic regression,in practical application,combined with the artificial rules and trading risk labels,setting the threshold value,Score and screen users' characteristics,and send SMS warnings or close their trading accounts to users who transact illegally,so as to reduce the risk of the bank.For the practical application scheme of the risk control model based on credit card transaction data,taking a large number of bank credit card transaction data as the experimental object,this paper has carried out the experimental data processing,user portrait feature modeling,model algorithm selection and implementation,and the construction of the risk control system.In the actual project application,our risk control model has achieved satisfactory results in capturing and preventing abnormal transaction behaviors of users in the "set integral" model and the "pseudo overseas" model,and the "abnormal Facebook system" developed by us has been widely used in the actual business of enterprises.