Presenting a comparative model: improving the quality of the internal financial information environment of government and non-government companies based on management characteristics (artificial neural network model and regression method

Subject and Purpose of the Article: 

The purpose of the current research was to present an optimal model for improving the quality of the internal information environment of companies accepted in the Tehran Stock Exchange based on the management features of government and non-government owned companies with the help of a comparative analysis of the artificial neural network model with the regression method.This study is applied and the method used is descriptive ex post facto (causal). To test the research hypotheses, a sample of 72 public companies and 38 nongovernmental organizations was selected during the years 1388-1400.We used panel data method to analyze the data and to test the first hypothesis, used the mean comparison test of two societies and used multiple linear regression and multilevel perceptron neural networks to predict the quality of the internal information environment.

Comparison of these two criteria showed that in addition to the effect of the management features of public and private companies on improving the quality of the internal information environment, in hypothesis 1-2 and hypothesis 2-2, with a coefficient of explanation of 0.67 and 0.75, respectively, compared to the MLP neural network method With a value of R2 0.47 and 0.50, it is preferred, but in the third hypothesis,The results showed that the regression method is more accurate in estimating and explaining the quality of the internal information environment of companies due to the greater number of observations compared to neural networks in this research.