Human Stress Detection in and through Sleep using Machine Learning

The effects of stress on the body and mind may be devastating, and it is a prevalent health problem. The quality of one's sleep is an important health indicator, and stress is a known disruptor of that quality. As a result, gauging a person's stress levels and health risks based on their sleep patterns may be quite informative. One's capacity for making decisions, paying attention, learning, and solving problems are all significantly impacted by stress. Both the disciplines of psychology and computer science have recently dedicated significant resources to the study of stress detection and modelling. Affective states, or the actual experience of having that emotion, are the measuring stick that psychologists use to assess the magnitude of stress. The majority of the progress in stress classification has been made using user-dependent models that cannot generalize to a new user. This results in substantial time investment from new users in training the algorithm to forecast their emotional states. Common mental health issues in children need to be addressed immediately since they may escalate into more serious issues if left untreated. The analysis and diagnosis of medical data are areas where machine learning techniques now excel. Different machine learning algorithms have had their accuracy across the chosen attribute set compared. Data on sleep-related factors, such as heart rate variability and sleep architecture, will be collected by wearable devices and analyzed using machine learning methods such support vector machine (SVM), decision tree (DT), MLP, and Ada Boost classifier. By analyzing a person's sleep patterns with the help of a machine learning model, researchers have shown they can effectively identify signs of stress in people.