The Financial Data Analysis System Based On Decision Tree ID3Algorithm For State Taxation

Currently, the tax system within the financial management system is being used asan administrative management platform, basically used to handle the daily work offinancial accounting, financial data is used in large centralized model, with the frequentuse of the system every day provincial bureau database Will produce large amounts ofbusiness data, such historical data usually used only as a summary query, dataavailability is not high, wasting a lot of resources, more important is not to provideeffective leadership in decision-making support.An analysis method of financial data which based on decision tree ID3algorithmfor state taxation, and detailed implementation process of the method are made in thisdissertation. This research goal is to use sophisticated data classification analysismethods to extract financial data from the critical tax information, through theintelligent processing, decision support information to the formation of leadership,helping leaders to better deployment of future work and improve efficiency.In response to this research, the decision tree algorithm to construct an classifier areused in this dissertation, the experimental platform on three public data sets and threesets of financial expenditure categories of data validate the accuracy of prediction andclassification. The subject of the decision tree classification algorithm used is the use oflarge amounts of data in the training set to create multiple classifier, and then use asimple scoring weight voting algorithm, selected the best classification results, for thelarge-scale experimental evaluation shows data set, the algorithm has goodclassification performance.