Analysis Of Credit Card User Value Based On RFMN Model Of Combination Weighting Method

As mobile payment methods become mainstream,credit card risks are increasing day by day.There are many models used by scholars to study credit card risk.To a certain extent,they can reduce credit card credit risk.However,when determining the weight of each influencing factor,scholars only consider the subjective weighting method or the objective weighting method alone.,Did not combine the two.As one of the important tools and methods to measure customer value,the RFM model has not been used by scholars to study credit card risks.This paper innovatively combines the combination weighting method and the RFM model to establish a new model,which reduces to a certain extent the adverse effects of using the subjective weighting method and the objective weighting method alone in evaluating the value of credit card users.Also achieved the purpose of evaluating the value of credit card users.In the stage of constructing the RFMN model based on the combined weighting method,since the subjective weighting method focuses on the intention of the decision maker(that is,the degree of importance the decision maker attaches to different indicators),it is highly subjective,while the objective weighting method focuses on objectiveness,but There may be situations where the weights are opposite to the actual weights.This article abandons the use of the subjective weighting method or the objective weighting method alone,and chooses to introduce the combined weighting method that combines the advantages of the subjective and objective weighting methods into the traditional RFM model.Combining with the research objects of this article,redefine the dimensions of the original model and add new dimensions,establish an RFMN model based on the combination weighting method,and transform the traditional RFM model into a model suitable for measuring the value of credit card users.The formula for calculating the comprehensive value score of credit card users is introduced.In the model application stage,in order to verify that the RFMN model based on the combination weighting method established in this paper can obtain the optimal clustering of the data,this paper conducts K-Means for the RFMN model of the combination weighting method,analytic hierarchy process,and entropy method.Cluster analysis,the corresponding clustering results were obtained,and the idea of single-factor analysis of variance was used for reference.The mean square between the groups was used to evaluate the differences in the comprehensive value scores of different types of credit card users,and the aggregation of the RFMN model based on the combination weighting method was obtained.This conclusion is that the class result has a better effect on the division of credit card user value.The clustering result of the RFMN model based on the combined weighting method is used as the presentation of the optimal clustering result of the sample data in this paper,and the user classification is obtained.Perform PCA principal component analysis on classified users,select user characteristics,and refer to factor score coefficients to obtain the characteristics of each type of user.Considering K-Means clustering results,PCA principal component analysis results,data statistical analysis conclusions and data preprocessing conclusions,the sample users in this article are divided into six categories,namely important value users,important maintenance users,important development users,and general Keep users,low-value users and risk users,give user portraits and make marketing suggestions.