Deep Neural Network Approach for Multi Class Stress Detection Through Heart Rate Variability

Abstract

Stress is an innate reaction to expectations, particularly when seen as detrimental or dangerous. Prolonged chronic stress elevates the likelihood of mental health problems, including anxiety, depression, and sleep disturbances. Heart rate variability (HRV) is a prevalent metric for assessing stress; it denotes the fluctuations in time intervals between heartbeats, in contrast to heart rate, which is an average value. This research investigates heart rate variability (HRV) as a biomarker for stress and presents a convolutional neural network (CNN)-based model for multi-class stress classification, distinguishing between no stress, interruption stress, and time pressure stress. The model, verified with the SWELL-KW dataset, attained exceptional accuracy, exceeding current methodologies. This study emphasizes the significance of HRV characteristics for stress identification using variance analysis.