Published Paper Details:

The categorization of sleep disorders has proven to be of main importance in the advancement of human quality of life, due to the fact that sleep apnea and other sleep disorders affect the general health of a person. Sleep stage analysis is inherently difficult and vulnerable to human errors, meaning that great need lies in the development of strong machine learning algorithms for an effective classification of sleep disorders. The paper discusses a comparative study of conventional machine learning algorithms with for classifying sleep disorders using the publicly available Sleep Health and Lifestyle Dataset that contains 400 rows and 13 columns representing various features related to sleep and daily activities. Several parameters of different MLAs are optimized in the paper using a genetic algorithm to improve classification performance. The performance of the k-nearest neighbors (KNN), support vector machine (SVM), decision tree (DT), random forest (RF), and logistic regression algorithms were tested alongside XGBoost for an all-around check on their correctness. Experimental results show that classification accuracy of with PCA showed 94.44%, 93.06%, 94.44%, 94.44%, 93.06%, and 94.44%, on applying PCA showed result 95.83%, 94.44%, 94.44%, 94.44%, 94.44%, and 94.44% respectively. The Logistic Regression achieved the highest classification accuracy of 95.83%, and its precision, recall and F1-score values on the testing data were 94.50%, 94.44% and 94.44%, respectively. The findings of the study increase the efficiency of the proposed algorithms in classifying sleep disorder. It contributes to the efforts that have been made in enhancing the accuracy of diagnostics. This integrated review declares the immense potential of machine learning in creating high-quality, automatic sleep health solutions, which serve as doorsteps to future innovation in sleep disorder management.