Single Channel EEG Based Binary Sleep and Wake Classification using Entropy Based Features

This paper proposes a novel method for binary sleep and Wake classification using entropy-based features extracted from a single-channel electroencephalogram (EEG). This study aims to improve the accuracy of sleep and Wake classification, which has several applications such as in sleep research, sleep tracking, diagnosis of sleep disorders, human performance assessment, human factors engineering. The proposed method is evaluated using the publicly available UCDDB dataset. Results show that the method achieved high classification accuracy, with the Ensemble subspace KNN classifier achieving the highest accuracy of 94.3%, followed by the fine KNN classifier with an accuracy of 92%. A significant improvement in performance can be attributed to the use of entropy-based features in the proposed method. Based on the promising results of this study, it is evident that the proposed method can be applied to sleep medicine for the classification of sleep stages, which can potentially lead to better diagnosis and treatment of sleep disorders.