Stress Prediction of Working Employees in Different Environments Using Machine Learning Techniques

Mental stress is a growing concern, especially among employees across sectors, affecting both well-being and job performance. Stressors like excessive workload, tight deadlines, and high employer expectations often lead to health issues such as anxiety, depression, cognitive decline, cardiovascular diseases, and suicidal thoughts. This study aims to predict employee stress using physiological and behavioral indicators. Real-time data was collected from employees in medical, teaching, and IT sectors through an online stress survey. Three machine learning models, Random Forest (RF), Support Vector Machine (SVM), and K-Nearest Neighbors (KNN), were utilized for prediction. Among the models, the Random Forest (RF) algorithm showed the best performance with a 92% accuracy rate. The study's findings highlight the effectiveness of real-time data and machine learning in stress prediction. A web-based application was also developed to help organizations monitor and manage employee well-being. This research demonstrates the potential of predictive analytics and machine learning to support stress management, encouraging proactive measures for a healthier work environment.