The purpose of this paper is to analyze the dynamics of taxation in the Iranian economy. Many economic systems that require modeling are complex, nonlinear, and have unknown parameters that make it difficult and unreliable to measure them in the form of mathematical and econometric models. In this paper, using coding under the adaptive fuzzy neural system approach - ANFIS - a model for dynamic tax analysis in an oil exporting economy during the period 1369 to 1395 was simulated. In this regard, the input values of the components of the central bank, the government and households, as well as the component of the role of tax policy in dynamic conditions on government revenues, budget balance and Iran's trade balance have been calculated. The results indicate that the output component of "Dynamic Tax Analysis in Iran" is exactly in the fifth level of system output, this means it is good. Experiments such as data error (RMSE) at each stage demonstrate the validity of the model and the very high accuracy of the neural-artificial network and fuzzy logic calculations compared to the research econometric models.

Keywords: Tax Dynamic Analysis, Small Open Economy, Adaptive Neuro-fuzzy Inference System (ANFIS)

Full-Text [PDF 1869 kb]   (800 Downloads)