Design And Application Of Customer Risk Management System Based On Data Mining

In this paper,a rural credit cooperatives credit audit department customer risk management system construction as an example,give prominence to mining system in data monitoring and management functions,in the existing credit approval,corporate loans,retail credit,industrial chain financing platform to provide information for large data on the basis of the establishment of a diversified with the rural credit cooperatives credit audit department itself based on the demand of data mining customer risk management system.The existing data mining based customer risk management system of illegal data mining BP neural network intelligent analysis of the mining object is designed for small scale data.However,a large number of violations of the accumulation of customer risk management,credit approval department of rural credit cooperatives supervision data analysis system of increasingly large scale,BP neural network intelligent analysis algorithm simple often because of the improper selection of initial points to the main credit monitoring index of mining results is not ideal.The face of massive credit records and evaluation index data,this paper does not use the single BP neural network mining of data mining,but the combination of K means clustering algorithm,the optimization process of BP neural network initial point selection,puts forward a kind of improved BP neural network algorithm based on K-means clustering K.In this paper,the improved algorithm using a K clustering algorithm to optimize BP neural network initial point,through the analysis of individual neurons in the neural network to determine the number of initial points;on the other hand,the establishment of customer risk management evaluation model by the improved BP neural network,so as to reduce the negative impact of human factors evaluation index the selection of the model.This paper implements the integration of real-time monitoring and management system through the establishment of credit business,set up a platform for the exchange and sharing of credit information;secondly,combining with the daily business process,the design theory of data mining in the region based on customer risk management system of data mining,the highlights of the real-time monitoring function module,customer risk the risk signal preprocessing module and credit warning mission management function module data flow demand analysis process,improve the function of the design process of three modules,three display module.In the credit audit department customer risk management system requirements analysis,the system through the design of Stream Data image to illustrate the data input,processing and output process of the three functional modules.System for the three function modules of the Stream Data diagram designed IPO table,third design the overall structure of the database and the IPO table designed for the entity attribute diagram,entity relationship diagram and the main database table.In the part of system implementation,the system focus on the display of borrowing customer basic information monitoring and monitoring real-time data,credit risk and credit approval function to realize data breaches of data near the credit risk credit business data flow,data mining,illegal credit business illegal data CIS and DRGS data analysis;finally,after the black box test the system timely detection and correction of the loan customer information overflow,concurrent file integrity response ability is not enough and other issues.