Machine Learning-Based Stress Detection in IT Employees: A Data-Driven Approach for Workplace Well-being

The rapid evolution of the Information Technology (IT) industry has brought about unprecedented challenges for its workforce, including heightened job demands, tight deadlines, and continuous technological advancements. This paper investigates stress detection in IT employees using survey data and machine learning. In machine learning, classification algorithms play a crucial role in sorting data into predefined classes. This study employs algorithms like logistic regression, decision trees, Naïve Bayes, Random Forest, SVM, and KNN to identify stress patterns. The results reveal varying accuracies among these algorithms. The findings not only unveil stress patterns but also explore influential stress factors, providing nuanced insights for IT workplaces. By anchoring the analysis in survey data, the research aims to offer actionable, machine learning-driven insights for proactive organizational interventions to enhance well-being and resilience. Among the various machine learning algorithms, Random Forest shows the most accurate results of 85%. This marks a pivotal step toward data-driven, personalized stress management in the IT sector, promoting a healthier and more productive work environment.