Explanation of Financial Variables Effective in Predicting Turnaround: An Artificial Intelligence Approach

The main aim of the research was to identify the financial variables that are effective in predicting turnaround of the listed companies in the Tehran Stock Exchange and to predict turnaround by using artificial intelligence method. For this purpose, the information of 173 Distress Companies that came out of distress and turnaround was extracted during 1383 to 1399. Artificial Intelligence approach was used to analyze the data. In this approach, by using Lars and Relief Feature Selection Algorithms, 10 out of 54 financial variables which were effective in turnaround of companies were identified and then, the Learning Algorithm of Support Vector Machine and Decision Tree were used to evaluate the accuracy of the results of the identified variables in predicting turnaround. The results showed that Lars Feature Selection Method and Vector Machine Algorithm Support have better performance in predicting the time to exit from distress as compared to the Relief Feature Selection Method and Decision Tree Algorithm. Also, regardless of feature selection methods, support vector learning machine has a higher predictive power as compared to decision tree.