Determination of work disability classification resulting from occupational accidents in the construction sector using artificial intelligence methods

Occupational accidents cause significant losses in employees' lives and create various challenges for employers. This thesis aims to predict work disability resulting from workplace accidents in advance and to identify the factors influencing this process. In this regard, analyses were conducted using Artificial Intelligence methods, specifically Logistic Regression and Random Forest models. The study utilized a dataset containing information related to the workplace, employees, and accidents. Among the identified factors, variables with both positive and negative impacts on the disability rate were determined. By understanding the influence of these factors on the process, the goal is to enhance predictability, prevent adverse outcomes following workplace accidents, and contribute to more efficient management of processes. This study has contributed to the development of predictive methodologies for workplace accidents and demonstrated that Artificial Intelligence methods can serve as a crucial tool in managing unforeseen circumstances. The high performance of the models lays the foundation for future studies in this field, contributing to more efficient and effective handling of processes despite the challenges posed by unknown and unpredictable situations in workplace accident management.