Σύγχρονες μέθοδοι μηχανικής μάθησης για την πρόβλεψη διαταραχών ύπνου
Modern machine learning methods for the prediction of sleep disorders
View/
Keywords
Διαταραχές ύπνου ; Παράγοντες τρόπου ζωής ; Μηχανική μάθηση ; Προγνωστική μοντελοποίηση ; Δείκτες υγείας ; Αλγόριθμοι ταξινόμησηςAbstract
Sleep disorders constitute a significant public health issue, as they affect both physical and mental well-being and are strongly associated with lifestyle and health-related factors. The objective of this study is to investigate the relationship between sleep disorders and lifestyle characteristics and to evaluate the performance of machine learning algorithms in predicting the presence of sleep disorders. For this purpose, the Sleep Health and Lifestyle Dataset was utilized, containing demographic, physiological and behavioral features such as sleep duration, stress level, body mass index (BMI), physical activity, blood pressure and occupation. After appropriate data preprocessing and exploratory data analysis, several classification algorithms were implemented, namely Logistic Regression, k-Nearest Neighbors (k-NN), Random Forest and XGBoost. The models were evaluated using weighted average performance metrics, including Accuracy, Precision, Recall and F1-score. The experimental results demonstrated that all classifiers achieved high predictive performance. Logistic Regression and k-NN yielded the best results, reaching an accuracy of 95%, while Random Forest and XGBoost achieved an accuracy of 92%. These findings indicate that lifestyle and health-related features can effectively contribute to the prediction of sleep disorders. In conclusion, the proposed approach highlights the potential of machine learning techniques as supportive tools for the early identification of sleep disorders and the analysis of the impact of lifestyle factors on sleep health.
Department
Σχολή Τεχνολογιών Πληροφορικής και Επικοινωνιών. Τμήμα ΠληροφορικήςNumber of pages
52Language
GreekCollections
Except where otherwise noted, this item's license is described as
Αναφορά Δημιουργού-Μη Εμπορική Χρήση-Όχι Παράγωγα Έργα 3.0 Ελλάδα
Αναφορά Δημιουργού-Μη Εμπορική Χρήση-Όχι Παράγωγα Έργα 3.0 Ελλάδα