Machine learning-based Sleep Stage Classification Using Wearable Technology
Date:
Abstract
Quality sleep is essential for cognitive function, physical health, and overall well-being, yet accurately tracking sleep outside of clinical settings remains a challenge. Traditional sleep studies, such as polysomnography, require expensive and inconvenient monitoring. This research explores the use of data science and deep learning to classify sleep stages—light, deep, and REM—using physiological data from smart wearable devices. By analyzing heart rate variability, movement patterns, and oxygen levels, we aim to provide a non-invasive, accessible alternative for sleep monitoring. Our findings indicate that wearable-based models can offer valuable insights into sleep quality, potentially empowering individuals to improve their sleep health and long-term well-being.