Classification of Mental State Using a Muse Headband and Machine Learning Algorithm

This paper describes an experimental study on the automatic classification of mental states using electroencephalogram (EEG) signals and a machine learning algorithm. We adopted the statistical and frequency domain features of EEG signals captured by Muse headband, which correspond to three classes of mental state that include relaxation, concentration, and neutral states. We experimented with different lightweight machine learning models to get an optimum classification. The paper includes descriptions of the Muse headband, feature extraction techniques, and machine learning algorithm development. We found the Random Forest and XGBoost algorithms achieved over 99% accuracy for the three classes. The study’s findings demonstrate that the Muse headband and XGBoost algorithm can predict three mental states with remarkably high levels of accuracy, providing promise for the development of novel methods to identify mental health disorders and the ability to detect complex mental states.