The Study of Cases Selection in Value-Added Tax-The cases of An Office of National Taxation Bureau of The Central Area, Ministry of Finance

• Main Authors: Chong-Si You, 游崇熙
• Other Authors: Jim-Chen Ko
• Format: Others
• Language: zh-TW
• Published: 2014
• Online Access: http://ndltd.ncl.edu.tw/handle/3vqk68

碩士 === 朝陽科技大學 === 財務金融系 === 102 === Our study data is selected from An official office of National Taxation Bureau of the Central Area, Ministry of Finance ROC which include being value-added tax (VAT) levied cases 187 and not being VAT levied cases 145. We set up 10 independent variables based on the possible combination through the variations of VAT report and Annual Business income tax report, also using Neural Network and Fuzzy Decision Tree methods to build up cases selection model. We expect these models not only help tax officers to choose those tax avoidance cases more efficiently, but also through N01 cases to identify those categorized metric into more significant cases for tax levy uses, besides, in comparison with Linguistic Variables rule settings of Fuzzy Decision Tree, Neural Network are not appropriated to build tax selection cases model, while rule settings in Fuzzy Decision Tree are more practicable in tax selection cases model.