Enhancing EEG-Based Sleep Stage Prediction Using Machine Learning Techniques

Sleep staging is crucial for identifying and treating sleep disorders, but the traditional method, which relies on manual classification by technicians, is time-consuming and subjective. Polysomnography (PSG) data, recording physiological signals during sleep, is essential for categorizing sleep stages. Analyzing sleep is vital for overall health and cognitive function. Machine learning is increasingly used to automate this process. This study analyzes EEG data from sleeping individuals in 30-s intervals using machine learning methods like AdaBoost, KNNs, and random forest. The random forest model achieved an accuracy of 86% in classifying sleep stages, highlighting the potential of machine learning in improving sleep analysis compared to manual scoring.