Employee Mental Workload Classification in Industrial Workplaces: A Machine Learning Approach

Employees at industrial workplaces are expected to produce labour of a certain standard. They are instructed to improve their quality of work, and this may take a toll on their mental health. Mental workload directly affects employees’ performance, productivity, and well-being. Therefore, this paper conducts a comparative study for the classification of mental workload where a mental workload dataset is subjected to four machine learning classification models-Naïve Bayes, Extreme Gradient Boosting, Support Vector Machine and K-Nearest Neighbour. Their performance is measured against the performance metrics-accuracy, precision, recall and f1-score. Before synthetic minority oversampling method Support Vector Machine performed the best with 90.41% accuracy and K-Nearest Neighbour performed the best with 98.61% accuracy after Synthetic Method of Oversampling Technique.