Stress Ocare : An advance IoMT based physiological data analysis for anxiety status prediction using cloud computing

Abstract In modern times individuals are facing an important social challenge in the form of stress. Combining sensor devices that capture physiological, and brain waves data, this study develops a machine learning technique using cloud computing to recognize stress in people in social contexts. In this paper, we are comparing several classifiers, including Random Forest, Support Vector Machine, k-nearest neighbor and AdaBoost, and also inventing a method that uses sensor data in day-to-day life. It detects stress levels with high accuracy. Our results show that by combining data from all the sensors, we are able to accurately differentiate between the stressful and normal situations of humans. In addition, this paper also evaluates the individual capabilities of each sensor modality and its applicability for stress detection in real-time situations. Methods: We have provided unique technology to incorporate sensor signals using cloud computing. It monitors the user’s sweat level, temperature, heart rate variation, and EEG under various motion estimations and also chooses the best model to detect the anxiety level based on the user’s motion conditions. Results: Evaluation of algorithms using sample data reveals an overall concern detection accuracy of 94% along with a significant reduction in false positives compared to the ultramodern techniques.