Stress analysis of EEG signal using K-means clustering and hybrid classifier

A significant factor that affects both mental and physical health is stress. Measuring accurately the stress level is very important. In this research, we approach measuring the stress using an EEG signal. There are many techniques to measure this stress; the most common ones are PET, ECG, EMG, and MRI, which help in measuring the stress levels of people. Another technique that is very useful and common, i.e., EEG, is mostly used to do such data analysis. EEG plays a very important role in detecting and correcting brain illness. EEG measures the electrical activity of the brain using small metal electrodes attached to the scalp. This activity results in wave lines of EEG signals. Here, we used a combination of the K-means algorithm and SVM-supervised machine learning algorithm to analyze the EEG data. K-means algorithm is used to divide and attempt to cluster the data while the SVM algorithm is used to classify the stress level based on these clusters. The main objective of this model is to improve the accuracy of stress detection by using the strength of both supervised and unsupervised machine learning algorithms. Using this hybrid approach, we attempt to provide a more precise and reliable method for detecting stress in real-time EEG data, with wide application in biomedical diagnosis, mental health monitoring, and many more.