Improved Sleep Stage Classification using Machine Learning with Data Augmentation

Accurate sleep stage classification is essential for diagnosing sleep disorders and understanding sleep patterns. Scarcity of desired data is the fundamental constrain in automated classification algorithms. In particular when it is required to deal bio medical signal, the accuracy is much needed. The current proposes a solution for this problem through a novel machine learning-based approach to enhance classification accuracy using Support Vector Machine (SVM) and Random Forest (RF) models, combined with data augmentation techniques such as time warping and signal permutation. The Sleep-EDF dataset was used for model training and evaluation, ensuring robust validation of the proposed methodology. Experimental results demonstrate that the proposed approach achieves a classification accuracy of 95%, surpassing traditional methods. The integration of data augmentation improves feature diversity, enhancing model generalization and performance. Additionally, the proposed framework enables real-time monitoring, making it suitable for practical applications in sleep health assessment. These findings highlight the potential of machine learning and augmentation techniques in advancing automated sleep stage classification.