Stress Detection While Doing Exam Using EEG with Machine Learning Techniques

This paper presents a comprehensive system design, architecture, and methodology for detecting the psychological states of individuals based on their emitted brainwaves while exam. The system utilizes the NeuroSky Mindwave Mobile device for EEG calibration and data acquisition. The obtained raw brainwave data are processed using Fast Fourier Transform to convert it into EEG brain electrical signal frequencies. A Convolutional Neural Network with three layers (input, hidden, and output) is employed for the classification task. The dataset consists of brainwave recordings from seven individuals, split into training and testing sets. The system achieves rapid convergence and optimal results after approximately 20 iterations, with a mean squared error (MSE) loss of 0.096 and a validation accuracy of 95.23%. The F1-score analysis demonstrates superior performance in predicting Class 1 (attention) compared to Class 0 for low stressed. These findings highlight the system’s ability to accurately detect the psychological states of individuals based on their brainwave patterns, particularly in predicting attention-related states.