Acute Stress State Classification Based on Electrodermal Activity Modeling

Acute stress is a physiological condition that may induce several neural dysfunctions with significant impact on life quality. Accordingly, it would be important to monitor in everyday unobtrusively and inexpensively. In this paper, we presented new methodological pipeline recognize acute conditions using electrodermal activity (EDA) exclusively. Particularly, combined rigorous robust model (cvxEDA) for EDA processing decomposition, an algorithm based support vector machine classify the state at single-subject level. Indeed, our method, single sensor, noise, applies phasic implements unbiased multiclass classification. To end, analyzed of 65 volunteers subjected different stimuli induced by modified version Trier Social Stress Test. Our results show successfully detected average accuracy 94.62 percent. Besides, proposed further 4-class pattern recognition system able distinguish between non-stress three stressful achieving as high 75.00 These results, obtained under controlled conditions, are first step towards applications ecological scenarios.