translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
PolySomnoGraphic (PSG) signals can be used to diagnose the disorders of sleep and to measure electrical brain activities. However, it is difficult to identify the behaviour of sleep by visual inspection. Thus, it requires an intelligent algorithm with computational techniques to identify the mild difficulties that occur in sleep. This paper will analyse PSG signals of both male and female subjects with normal sleep patterns and mild difficulties in sleeping by extracting Power Spectral Density (PSD) features. The dimensionality of extracted features reduced using Linear Discriminant Analysis (LDA) to study the best feature selection for the SVM classifier. The comparison results of each classification task which is defined as a classification of normal and mild difficulty PSG signals; including both male and female subjects, classification of both normal and mild difficulty PSG signals in male subjects only, and classification of both normal and mild difficulty PSG signals in female subjects only are presented. The computed result shows 87.5%, 50% and 75% classification accuracies for each classification task, respectively, with higher sensitivity and specificity rates.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.