Validation of Artificial Intelligence Algorithms in Predicting Financial Distress in the Industrial and Mining Sector with Emphasis on the Role of Macroeconomic, Financial, Managerial and Risk

Predicting financial distress is an important phenomenon for investors, creditors and other users of financial information. Determining the probability of a company’s distress before occurrence of distress and bankruptcy is considered a very interesting and attractive subject and can be useful for both managers, and investors and creditors. In this study, using the data of 1350 year- company during the period 2008 to 2016 in industry and mining sector in Iran, the factors affecting financial distress and predicting it through Intelligence Algorithms methods (decision tree, support vector machine, and Bayes classification methods) have been studied. The results of the study indicate direct impact of financial risk and inflation, and inverse` impact of the ratio of non-executive directors, stock returns, and the ratio of operating cash flow on financial distress. The results also show that decision tree method, using financial and economic data, has higher efficiency in predicting financial distress compared to Bayes classification method and support vector machine.