HYBRID ENSEMBLE VOTING MODEL FOR ENHANCED SLEEP DISORDER DIAGNOSIS

Abstract

Since sleep-related problems have a substantial impact on people's health and quality of life, accurate classification of sleep disorders is crucial for efficient diagnosis and treatment. Even though individual deep learning and machine learning models have shown encouraging outcomes, model bias and overfitting frequently limit their performance. This study offers an ensemble learning-based extension for the categorization of sleep disorders that integrates predictions from several machine learning models in order to overcome these difficulties. To increase robustness and classification accuracy, the outputs of optimized base classifiers are aggregated using a voting classifier. Experiments on the Sleep Health and Lifestyle Dataset show that the suggested ensemble strategy outperforms individual models with an accuracy of 97.3%. Additionally, to improve system usability and real-time user engagement, a Flask-based web interface with secure user authentication is created. The suggested extension offers an automated framework for diagnosing sleep disorders that is dependable, accurate, and easy to use.