The Application Of Data Mining Algorithm Intax Source Management Under Large Data Background

"Big data" has become one of the popular new words,large business data,medical data,large government data.How can we use the data and related technical means to distinguish and filter different types of tax sources in the background of large data,which can provide great reference for tax source management and provide the problems that need to be faced and solved for strengthening tax source management.Decision reference.In this paper,the K-means data mining technology of large data analysis is used to analyze the financial data of the manufacturing companies listed in Shanghai stock market.From the correlation analysis,after the analysis of principal components,both analyzes are used to remove the data The relative value of the relatively independent variables,screening is conducive to the cluster analysis of financial data,and ultimately get the corporate income tax,operating cash flow net,total assets,tax payable four dimension variables.These four variables are less suitable for clustering analysis.In the actual cluster analysis,the results can be seen on the 433 listed companies are accurately divided into three categories,almost no coincidence with each other,the fitting effect is better.However,the results of the specific classification of the results,far from being interpreted.For the effective interpretation of clustering results,in order to more reasonable service in the tax source management.This paper consists of six parts:The first part is the introduction,through the current large data background data mining technology in the application of tax source management analysis,proposed clustering analysis applied to the concept of tax source management.The second part introduces the concept of data mining and the principle of cluster analysis,and makes a theoretical basis for the follow-up writing.The third part introduces the current situation and the role of tax source managementIn the fourth part,the correlation analysis,principal component analysis and cluster analysis are carried out,and the practical application of data mining technology applied to tax source management is put forward.The fifth part is based on the fourth part of the variables obtained on the basis of further discussion,respectively,from the financial perspective of corporate income tax,operating cash flow net,total assets,tax payable four dimensions variable analysis of the source management methods As well as possible problems.The sixth part is the conclusion.The main conclusions are as follows: First,find a way to regulate corporate taxation;Second,simplify the tax source management indicators;Third,to provide a source of tax management ideas;Fourth,the establishment of a sound tax database Very necessary.One possible innovation in this paper is that the purpose of clustering analysis can be achieved by selecting less financial data,eliminating the need to study cumbersome financial data.Another possible innovation is to find a way to regulate corporate taxation,both the target enterprises to join the sample set,by checking the abnormal value of the way to determine whether the normal tax business.