Empirical stress prediction among drivers utilizing wearable sensors and psychometric signals: Towards smart health

Abstract A psychophysiological state i.e. mental stress has an inauspicious effect on the quality of human life. To ameliorate mental and physical health, accurate detection of mental state can be propitious to deliver prevention and treatment for this disease. Subsequently, unobtrusive monitoring of physiological signals is made possible by deploying wearables in everyday life. In this context, this paper proposes a stress detection model by incorporating wearable sensors for detection purposes. Three vital physiological sensors i.e. blood pressure, heart rate, and respiration rate are utilized for stress detection in this study. A methodology supporting five classifiers, capable of predicting cognitive degradation in performance is proposed to classify mental stress into five stages. The IBK classifier outperforms all the other classifiers in terms of accuracy, precision, f-measure, and recall. The ability of this model to detect mental stress unobtrusively can help to prevent the rate of accidents and productivity losses on roads.