Design And Implementation Of Data Mining Based Finical Data Analysis System

With the popularities of finical management information system,a mass of data are accumulated.But the mass data have not been effectively used,and manager of finical facility can only acquire some superficial information by some simple statistics and sort method of statement.This finical data are acquired by management and history accumulate,and it may contain many useful information.How to find this information using deeply data mining methods,so as to prevent and control the finical risk is a problem that is needed to be intensive studied.A finical data analysis system based on data mining methods is put forward in this thesis to solve this problem.The study includes the content as below:1)Design the model of finical data analysis system.The model includes 5 function modules: data collection,training,storage of the rule,analyze and inference and application and presentation.The analysis procedure includes data mining methods and traditional case matched methods,and the inference module are implemented by the combination of utilization of data mining methods and case matched methods.2)The construction of finical data warehouse.The finical data are derived from two ways: internal finical facilities and out of finical facilities.The structure of the finical data warehouse is put forward in this thesis;the star schema and metadata are studied.3)Data mining method based on random decision tree is present.The random decision tree is used,and the experiments are carried out to verify the effectiveness of the method.4)The traditional case match methods are improved.The traditional case match method based on distance is improved with the using of cluster method into the case match procedure,and is can improve the efficiency and accuracy of the case speculative.5)A finical data analysis system is designed and partially implemented for a bank in Han Dan based on the methods and model in this dissertation.