Machine learning-based signal processing using physiological signals for stress detection

Abstract

Stress is a common part of daily life which most people struggle in different occasions. However, having stress for a long time, or a high level of stress will jeopardize our safety, and will disrupt our normal life. Consequently, performance and management ability in critical situations degrade significantly. Therefore, it is necessary to have information in stress cognition and design systems with the ability of stress cognition. In this paper a signal processing approach is introduced based on machine learning algorithms. We used collected biological data such as Respiration, GSR Hand, GSR Foot, Heart Rate and EMG, from different subjects in different situations and places, while they were driving. Then, data segmentation for various time intervals such 100, 200 and 300 seconds is performed for different stress level. We extracted statistical features from the segmented data, and feed this features to the available classifier. We used KNN, K-nearest neighbor, and support vector machine which are the most common classifiers. We classified the stress into three levels: low, medium, and high. Our results show that the stress level can be detected by accuracy of 98.41% for 100 seconds and 200 seconds time intervals and 99% for 300 seconds time intervals.