The Application of Artificial Neural Network in Predicting Future Cash Flows

Document Type : Original Article

Authors

1 Professor of Accounting, University of Allameh Tabatabai, Tehran, Iran

2 Assistant of Professor of Accounting, Institute of Higher Education Raja, Tehran, Iran

3 M.A. in Auditing, Institute of Higher Education Raja, Tehran, Iran

Abstract

Cash flow of resources is essential for any economic unit and creates a balance between available cash and cash needs, the most important factor is the economic health of that unit. Since the liquidity situation is based on the judgment of many interested parties such as shareholders and investors about the position of the economic unit, therefore, predicting future cash flow is of utmost importance.In addition, providing the appropriate model to predict accurately with minimal deviation of accounting knowledge has been of interest to many researchers. The purpose of this research was to use neural network and multilayer perceptron and determine the best model by using commitment regression model to predict cash flow. For this purpose, two hundred and eighty seven companies in Tehran stock Exchange were investigated during 2001-2011. The results of the review of various neural network models suggest that the two structures with 8 and 11 hidden nodes, is the best model to predict cash flow.

Keywords

Journal of Iranian Accounting Review
Volume 3, Issue 9
April 2016
Pages 63-80

Files

Share

How to cite

Statistics