sleep disorders, particularly sleep apnea, have an impact on people’s health, and that reveals the need for a correct diagnosis. However, sleep experts use complex and time-consuming methods for the manual classification of the various sleep stages. Based on the Sleep Disorder Data which is available for public access and consists of 400 records and 13 attributes, this work introduces a machine learning classification model. A number of deep and techniques-based machine learning models are considered and their performance is evaluated in accurately diagnosing sleep disorders. Lifestyle parameters and sleep health characteristics are among the features in the dataset meaningful for discovering patterns, with patterns of which can be indicative of existing sleep-related disorders. Based on the models assessed, it was found that the model with the highest performances are bagged models particularly the Voting Classifier with RF and DT. The accuracy, precision, recall, and F1-score of the algorithm were 0.973, suggesting that the algorithm useful for sleep disorder classification and is reliable. These results imply that the proposed machine learning approaches provides an opportunity to make smarter, faster and more accurate sleep disorders diagnoses in order to enhance the possibilities of the physicians’ decision-making process and patients’ condition.
Kartheek et al. (Tue,) studied this question.