The Competition between Regression and Artificial Neural Network Models in Earning Management Prediction

The Competition between Regression and Artificial Neural Network Models in Earning Management Prediction

International Review of Business Research Papers

Vol. 10. No. 2. , September 2014, Pages: 148 – 159

The Competition between Regression and Artificial Neural Network Models in Earning Management Prediction

Hossein Etemadi and Hiva Rastegar Moghadam

Earning management prediction is one of the salient issues that help stockholders and other users in accurate evaluation of financial position and performance of the firm and it becomes more and more important in preparing and reporting the useful information for decision makings regarding the highlighted importance of recent worldwide financial crisis. In this paper, researchers attempt to provide an optimum model for predicting the earning management.The main goal of this paper is to compare the power of artificial neural network model with the multiple linear regression one in earning management prediction. The sample pool consists of 107 firms in Tehran Stock Exchange (TSE) for the time period 2002 - 2010. The results of this study show that artificial neural network is more powerful than multiple-linear regression in predicting earning management.

Field of Research: Accounting

DOI : http://dx.doi.org/10.21102/irbrp.2014.09.102.11