Prediction of Earnings Management by Use of Multilayer Perceptron Neural Networks with Two Hidden Layers in Various Industries
Over the last few years, increase in the number of financial crisis cases has attracted the attention of a great number of investors and creditors to the prediction of these cases by using the data available from financial statements of the corporations in order to prevent losses which are caused by them. Many of the studies about the earning management only focus on identifying the factors which affect the earnings management and relation between earnings management and these factors and not on using these factors in order to predict the earnings management. On the other hand, in the present study, 12 influential factors, which are introduced by former studies, are used to evaluate the predicting ability of the neural networks for prediction of earnings management and its different levels. Multilayer perceptron neural network with 2 hidden layers is used for prediction. Moreover, the required data is obtained from the financial statements of the listed companies of Tehran Stock Exchange (TSE). The results show that the applied neural network method has an acceptable ability for the prediction of earnings management and its different levels in all different studies industries.
Keywords: Earnings management, Neural network, Multilayer perceptron, Accrual items
Reference to this paper should be made as follows: Mahmoudi, S., Mahmoudi, S., & Mahmoudi, A. (2017). “Prediction of Earnings Management by Use of Multilayer Perceptron Neural Networks with Two Hidden Layers in Various Industries”, Journal of Entrepreneurship, Business and Economics, Vol. 5, No. 1, pp. 216–236.