Abstract Occupational Health and Safety (OHS), which holds substantial importance due to its profound human implications and far-reaching economic consequences, requires effective management to minimize workplace accidents and occupational diseases through data-driven risk assessment and proactive safety strategies. Achieving this necessitates the availability of historical data and its examination through appropriate analytical methodologies, which is critically important for determining evidence-based preventive measures. Thus, this study analyses Turkey’s national OHS records from 2010 to 2022 using statistical analysis and machine learning approaches to evaluate sectoral risks, predict accident trends, and identify key determinants of workplace hazards. Results indicate substantial sector differences, with coal mining and heavy industry showing the highest accident and fatality rates. Machine learning models demonstrate strong predictive capability, with gradient boosting providing the best work accident prediction performance and random forests achieving the best performance for occupational disease prediction. Clustering analysis identifies three distinct industrial risk groups, while Principal Component Analysis (PCA) reveals regional disparities, particularly in highly industrialized provinces such as Istanbul, Kocaeli, and Izmir. Classification models further achieve over 98% accuracy in identifying high-risk groups, highlighting the potential of machine learning for proactive OHS management. The findings of this paper provide actionable insights for policymakers and industry leaders to optimize safety regulations and develop targeted interventions and strategies to reduce workplace risks and identify which sectors and worker groups should be prioritized.
Aksoy et al. (Sat,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: