Design And Implementation Of The Bank Customer Risk Prediction System

Due to the uncertainties of banks' financial services and clients,there are too many uncertainties in the bank's future operation and high risks.The good condition of bank loans depends on the good condition of customer resources.The financial informationization is conducive to the automatic monitoring,assessment and early warning of risk factors,reducing the risk of occurrence of risks and safeguarding the normal order of the financial industry.With the rapid development of the Internet,the bank's operating system has lagged behind the current economic development.It is urgent to reform the management model of outdated business customers,strengthen the monitoring and management of credit risk,and financial risk based on big data Control is the use of set data analysis model and early warning mechanism to automatically analyze and early warning of customer credit risk.By establishing risk information database management,unified management of risk data transmission is realized,risk data is shared within the system,waste of resources is reduced,and risk control pre-control is enhanced.In this paper,we use the literature research method,structured method,case analysis method and comparative analysis method to analyze and study the large bank of customers,using the current big data mining,cloud computing and other advanced computer technologies,combined with the mature Apriori algorithm data mining analysis Method to digitally analyze bank customers' risks.Apriori algorithm is one of the most classic and most commonly used algorithms in data mining.However,bank client data belongs to mass-scale data,and the computation of data has risen sharply.In this paper,B / S-based J2 EE architecture will be used to store data using Hadoop database.On the basis of this,Apriori algorithm in bank risk association rules will be improved.The idea of Apriori algorithm improvement is to merge upper triangular support matrix with transaction binary vectors,so as to improve the speed of generating frequent 2-itemsets,so as to improve the execution efficiency of the algorithm and at the same time guarantee the strong correlation between the attributes of customer risk factors The validity of the rules.According to the actual situation,the improved algorithm has better execution efficiency than the traditional Apriori algorithm under the condition of the same set of experimental data,different number of transactions,different support degrees,different number of items and so on.According to business needs,this paper reconstructs the bank customer risk assessment process under traditional manual mode,analyzes each business process in detail and divides thesystem into customer information management,data information management,data processing,customer risk analysis,customer credit forecasting,System management and other six functional modules.This paper establishes a use case diagram of each function module through UML,elaborates each use case situation,carries on the overall design to the system architecture system and the function module,and designs the class diagram and database of the system function module in detail.Through the Java programming language development Realize each function module.In the system implementation,aiming at the potential non-performing loans of banks and related risk clients,this paper uses the Apriori algorithm based on the improvement of large data customers from the analysis of risk factors and their relevance,and then concluded that certain rules and conclusions,the establishment of Risk early warning model,which is the most important result of this thesis.The system test results show that this paper develops a bank customer risk prediction system based on Hadoop database and improved Apriori algorithm to quickly extract the customer risk situation,and the similarities with the offline manual audit results of more than 95%,fully able to assist bank staff to customer risk Forecast,while using big data analysis method with high reliability,high efficiency and reduce the factors of man-made credit,to improve the efficiency of bank credit business is of great significance.