Selection for value added tax audit

In this article in order manage the risk from the lack of tax compliance.The optimal model “selection forvat audit” using data mining techniques  (vector machine and decision tree) be studied. In the selection for auditing, the important issue is select the right files for auditing from within the range of potential candidates .So that’s the choice while covering the maximum tax risk, it will increase the efficiency of the tax system. because of the limits available in time and sources, Audit all tax payer will get various costs . In addition, the costs associated with the audit and the revenue from auditing are also raised .therefore, the output of the tax audit process greatly contributes depending on the quality of the selection of the audited tax payers collection. To do this research in it all, using the Delphi method, factors affecting tax risk are identified and continuing with the using of the statistical samples of research. Which includes 281 companies that were subject to Vat audit between 2001 to 2015,Different models were extracted and their ability  was evaluated to choose optimally for Vat audit. The results of the study show that the Linear support vector machine method up to 70% of tax payers with this clustering algorithm correctly they will be separated depending on the analysis of existing data. Which is more  reliaboh le then other data mining techniques of decision tree.