Data Mining Technology In The Data Centers Of Large Commercial Banks

With the finish of centralizing financial data stored in data center of commercial bank,every commercial bank will challenge the difficulty on how to extract the valuable information forming the knowledge which can be used to improve the competitive capacity by providing the decisions support to enterprise.The data mining technology make it possible that the data center of commercial bank becomes the knowledge base.Now it is imminently important that the commercial bank use the data stored in data center to classify their customers,analyze the transactions correlation,and therefore guarantee the reliable services by utilizing the Data Mining and Data Warehouse technique.By classifying their customers,bank can get well understand the markets' demands,then provide specific financial products for different customer groups according to their preferences,and be trusted by all their customers.By analyzing transcations' inherent correlation,the bank can get the more effective and trustworthy indirect information than ever which can be used to support business operation and establish information technology plan.By forecasting the transaction association in the future,the bank can provide the highest performance service level with the lowest cost, and avoid the service outage effected by the problem of computer system's capacity.Firstly,this thesis introduced the concept of data mining and data warehouse,studied the technical foundation of data mining and the principle concerning and finished the mining tool;then combined the bank of data,designed the model of the data mining management system and data warehouse;lastly implemented the model by finishing the association rule mining algorithms,clustering algorithms,regression analysis algorithms.This thesis mainly includes next four contributions:(1)Designed the data warehouse which integrated various information of bank,including transaction log,customer's information and control information.By filtering,cleaning and integrating these information, the data warehouse effectively support the data mining.(2)Mined lots of association rules from transaction log,which can be used to reorganize the existed the transaction,and lastly provided special business for customers.(3)Applied the clustering algorithm to classify customers according the difference of their personality,income,and favorite, then provided the customer group with appropriate service,which improve the bank's competitive capacity greatly.(4)Set up the model of relationship between quantity of transaction and usage of CPU by applying regression analysis algorithms.According to this model,the upper limit transactions can be conveniently estimated with current CPU capacity.Which prevent the possible capacity risk in advance and provide the support for optimization and extension of capacity.