Toward a socially sustainable workforce in service operations: Predicting employee burnout with ensemble machine learning

Employee burnout, characterized by emotional exhaustion, depersonalization, and low accomplishment, poses a significant threat to social sustainability within service organizations. It not only impacts employee well-being but also undermines workforce stability and organizational productivity. This study explores the efficacy of ensemble machine learning models in predicting burnout risk, contributing to a more socially sustainable workforce. We implemented a bagging ensemble technique, training a diverse set of base learners, including decision trees, support vector regressors, and random forests, on distinct data subsets. Feature importance analysis was employed to identify the most influential features contributing to burnout. These features included workload allocation, as measured by resource allocation metrics, mental fatigue scores reported by employees, and tenure. Hyperparameter tuning was conducted using a grid search approach to identify the optimal configuration for each base learner within the ensemble. The developed ensemble model achieved a high R-squared value of 0.9567, demonstrating its effectiveness in predicting employee burnout. Analysis of feature importance revealed critical insights into the underlying factors associated with burnout. These insights can inform the development of targeted organizational interventions aimed at mitigating burnout risk and promoting workforce sustainability. Policy recommendations include bolstering mental health support, promoting work-life balance through flexible arrangements, optimizing workload distribution, implementing career development initiatives, and cultivating a supportive work environment. Exploring alternative ensemble methods like boosting and stacking, investigating the potential of Automated Machine Learning for efficient model development, and utilizing unsupervised learning techniques for employee risk segmentation are promising avenues for further exploration.