Sleep Disorder Classification with Ensemble Based Machine Learning Models

Abstract

Sleep disorders, in particular sleep apnea, have a negative impact on the health of people, so it is important to diagnose them correctly. However, the manual classification of the individual sleep stages is complex and labor intensive in nature and is used by sleep experts. In this paper, a machine learning classification model based on publicly available Sleep Disorder Data, which contains 400 records and 13 attributes, is presented. Various deep learning and technique-driven machine learning models are reviewed and their effectiveness in accurately detecting sleep disorders is evaluated. The variables of lifestyle measures and sleep health measurements are of great importance in the dataset to determine the trends that can be used to suggest the presence of sleep-related conditions. The analysis of the models showed that the best-performing models are bagged models, namely, the Voting Classifier with the help of the Random Forest and Decision Tree algorithms. The algorithm accuracy, precision, recall, and F1-score were 0.973, which means that it can be used to categorize sleep disorders and is reliable. The results indicate that the suggested machine learning methods have the potential for more intelligent, rapid, and precise diagnostics of sleep disorders, hence improving physicians' decision making and patients' conditions.