A New Approach for Sleeping Disorder Classification using Supervised Machine Learning Algorithms

Sleep disorders are now a prevalent health condition worldwide and are identified as contributory causes of several chronic diseases like cardiovascular diseases, obesity, and reduced intellectual functions. Intemperate rates of sleep disorders such as insomnia and sleep apnea are shown to be related to abnormal sleeping hours, unsatisfactory lifestyles, and demographic variables. Here, we propose a machine learning method to predict sleep disorders with high accuracy on the Sleep Health and Lifestyle dataset, using key features like age, gender, BMI, physical activity, occupational type, sleep duration, and quality. We employ four supervised machine learning models: Logistic Regression, XGBoost, Decision Tree (DT), and Support Vector Machine (SVM). Data preprocessing techniques such as handling missing values, normalization, and label encoding are conducted to ensure data quality. Feature importance analysis is also performed to identify influential attributes such as BMI, blood pressure, and sleep quality in the prediction of sleep disorders. The measurements of precision, accuracy, recall, and F1-score are utilized to evaluate the models' performance. Hyperparameter tuning techniques are used for maximizing model performance. Our results indicate that these machine learning models are capable of classifying sleep disorders with high accuracy, with XGBoost and SVM demonstrating greater prediction accuracy. This study shows how machine learning technology can be utilized in medicine to analyze and treat sleep disorders early, helping medical professionals make better decisions and enhancing patient outcomes. The addition of genetic algorithm-based optimization methods further boosts classification accuracy by optimizing the parameters of the model. Comparative studies find the dominance of ensemble methods over traditional classifiers in solving high-dimensional healthcare data. Additionally, the system's ability to include variable lifestyle-related factors presents an integrated solution for individual healthcare. Such findings unveil pathways for future research in designing effective and automated diagnostic systems for sleep disorders.