Use of Biomedical Signal for Human Emotion & Stress Analysis Using Machine Learning Approach
Automatic recognition of brain activity and stress is critical for ecologically friendly explanations. A bad day in the mind and stress arena has become a significant source of worry for our society, contributing to many health problems and enormous business losses. Often, stress is defined as the body’s response to professed physical, psychological, or expressive pain. To effectively address and resolve these issues, a completely automated solution is necessary. With this purpose, this study attempts to provide a framework for a more accurate and reliable stress identification system based on electroencephalography (EEG) data. It extracts the critical characteristics for stress identification via several frequency bands and principal component analysis (kernel PCA). This procedure converts the EEG indication to a more edifying signal that may be used to generate sample records. It provides extracted data from that sample dataset to elicit the desired level of tension in the individual. Three EEG frequency bands were employed in this study together with a machine learning technique to extract crucial data in order to improve stress detection classification accuracy.