Analyzing and predicting taxation rate with emphasis on the role of accounting measures and the governance system of the artificial intelligence approach of choosing the variable of neighborhood analysis and hyperdisk (NCA-HDLMC)

The purpose of this study is to analyze the factors affecting the taxation rate on companies listed on the Tehran Stock Exchange. This analysis was performed using the Neighborhood Analysis Algorithm and the Hyperdisk in a linear and nonlinear manner. The initial independent variables in this study include the variables of the governance system and the variables of accounting. To measure the taxation rate, the taxes rate paid by the company has been used. Experimental findings related to the study of 150 companies listed on the Tehran Stock Exchange in the period from 2012 to 2019, show that using the neighborhood analysis algorithm as a method of switching accounting criteria (quick ratio, return on assets, sales returns, ratio of sales to total assets) have the greatest impact on the company s taxation rate. Other research findings also showed that the nonlinear algorithm has a higher power than the linear algorithm in predicting taxation.