Predicting Firm Profitability Using Data Mining in the GCC Region

Firms' financial health is essential for all stakeholders because it directly affects economic stability and may lead to insolvency issues impacting national economies. Technological advancements and the rise of Big Data (BD) have created an increasing necessity for effective methods to uncover hidden patterns in Gulf Cooperation Council (GCC) firms' data, presented in Data Mining (DM). This study aims to evaluate the accuracy of DM techniques in predicting the profitability of GCC-listed firms. The methodology utilized in this study follows the Cross-Industry Standard Process for Data Mining (CRISP-DM), which includes the phases of business understanding, data understanding, data preparation, modeling, and evaluation. The datasets were obtained from the London Stock Exchange Group (LSEG) and the World Development Indicators (WDI). Employing seven DM models, single and ensemble, the study predicts firm profitability presented in Return on Investment (ROI). The results indicate that CatBoost outperformed all other models, achieving the highest R-squared value with 0.4684 and the lowest error rates in ROI prediction. In contrast, linear regression emerged as the weakest performer. The study provides important insights into the application of DM techniques for firm performance prediction within the GCC region, offering a foundation for future advancements in the field.