Research on Tax Inspection Case Selection Model Based on Bayesian Network

Abstract

Case selection is the first procedure of inspection, and the accuracy of case selection is related to the quality of the whole inspection work.The accuracy of case selection relates to the quality of the whole audit work. Through the selection of financial indicators of enterprises, the Bayesian network method is applied to the field of tax inspection and case selection. With the help of the reasoning function of the network, the probability of occurrence is predicted, and a case selection model is established to reflect the uncertainty and complexity of the case selection system, so as to realize the prediction and diagnosis of the case selection system. Taking the honesty state of the selected enterprise as a probability event, the accuracy of this model is 93.3%. It shows that the model can not only improve the accuracy of tax audit case selection, but also provide the method guidance and decision-making reference for tax audit case selection.