نوع مقاله : مقاله پژوهشی
عنوان مقاله English
نویسندگان English
The purpose of this study is to develop a predictive modeling of bankruptcy, focusing on modern measurement methods such as Neural Networks and Support Vector Machines, specifically for listed companies in the Tehran Stock Exchange Market. The findings of this research indicate that the utilization of new structures, such as hybrid intelligent systems based on data mining models, holds significant potential in detecting company bankruptcies at the national level. It should be noted that the data used in this research was generalized to two industries, namely food and textile, and was not limited to data solely from stock-exchange-listed companies. It is observed that in the two industries under study, the financial ratio of "Accumulated Profits to Total Assets" has the highest recurrence in predicting bankruptcies. For this reason, the aforementioned ratio will be chosen as the top ratio for predicting bankruptcy. The results reveal another important issue. It is observed that in each industry, relatively different financial ratios have been selected as the financial ratios with the highest recurrence. This demonstrates that the possibility of considering a specific set of financial ratios as model inputs, as commonly done in most studies, is not valid. In each industry, the combination of inputs varies based on its characteristics and structure. Therefore, the findings of this section can be considered one of the important reasons for the superiority of the designed system in this research compared to various structures presented in domestic and international studies.
کلیدواژهها English