International Research Journal of Education and Technology

Sleep is a fundamental aspect of human physiology, with a profound impact on overall health and well- being. Sleep disorders, affecting a significant portion of the global population, present a pressing health challenge. To address this issue, our project aspires to develop a holistic methodology for sleep stage classification using EEG signals. This methodology holds the potential to revolutionize sleep science by enhancing our understanding of sleep physiology, improving sleep disorder diagnosis, and enabling personalized sleep therapy. Furthermore, it aims to contribute to sustainable and environmentally conscious sleep monitoring and therapy practices by optimizing energy consumption and resource utilization in sleep monitoring devices. The primary objective of this project is to advance the field of sleep medicine by accurately classifying sleep stages from EEG signals. Sleep is conventionally categorized into two main stages: Non-Rapid Eye Movement (NREM) and Rapid Eye Movement (REM). Accurate classification of these stages is essential for understanding sleep patterns and diagnosing sleep disorders. Leveraging sophisticated machine learning techniques, including Gaussian Mixture Models, K-Nearest Neighbors, Softmax Discriminant classifiers, and Logistic Regression, we aim to precisely delineate these sleep stages. Our project unfolds in several phases. We begin by collecting raw EEG data, a crucial prerequisite for accurate analysis. Data preprocessing techniques are then employed, involving noise reduction, filtering, and artifact removal, to enhance data quality.

Keywords: REM, NREM, artifact removal, noise reduction, holistic methodology.