The Comparison of Multi-variable Linear Regression and Artificial Neutral Networks in Tax Evasion of Legal Persons in Iranian Tax System

Tax evasion is one of the most important problems of tax system in the most countries around the world. It covers any unlawful attempt to avoid paying taxes. In present study, the affective factors on tax evasion based on experts’ views were extracted by using Delphi method, so we identified 29 factors and finally 16 factors were extracted based on measurement ability among them. The statistical population of this study was the companies who had files in Tax organization. Based on Morgan’s table, 400 companies were selected as sample for performance of year 2012. The extracted data were analyzed based on multi-variable linear regression and artificial neutral networks, that both methods represent the effect of identified factors on tax evasion of companies. Then we analyzed the efficiency of multi-variable linear regression and artificial neutral networks, the results showed that artificial neutral networks have more efficiency in comparison with multi-variable linear regression. So, efficiency of multi-variable linear regression to detect tax evasion of legal persons was 60%, while efficiency of artificial neutral networks was 82.5%.