TY - GEN
T1 - A study on automatic sleep staging algorithm based on improved SVM
AU - Zhang, Song
AU - Hu, Bin
AU - Zheng, Xiangwei
N1 - Publisher Copyright:
© 2017 Association for Computing Machinery.
PY - 2017/9/22
Y1 - 2017/9/22
N2 - With the progress of science and technology and social development, public health becomes the focus of social concern. The study of sleep staging is an inevitable trend in the development of medical technology and the results can be used as an effective means of adjuvant therapy for some diseases such as insomnia, epilepsy, anxiety and so on. After analyzing least squares support vector machine (LSSVM) and decision tree, this paper proposes an improved SVM based on tree structure (DLSVM), which is applied to automatic sleep staging. After preprocessing the main components of the EEG signal waves at various stages of sleep, DLSVM applies different feature subsets to automatically classify the sleep stage. The simulation experiments show that DLSVM can reach 86.47% overall accuracy of sleep staging and is superior to other similar related algorithm.
AB - With the progress of science and technology and social development, public health becomes the focus of social concern. The study of sleep staging is an inevitable trend in the development of medical technology and the results can be used as an effective means of adjuvant therapy for some diseases such as insomnia, epilepsy, anxiety and so on. After analyzing least squares support vector machine (LSSVM) and decision tree, this paper proposes an improved SVM based on tree structure (DLSVM), which is applied to automatic sleep staging. After preprocessing the main components of the EEG signal waves at various stages of sleep, DLSVM applies different feature subsets to automatically classify the sleep stage. The simulation experiments show that DLSVM can reach 86.47% overall accuracy of sleep staging and is superior to other similar related algorithm.
KW - Automatic sleep staging
KW - Decision tree
KW - Feature extraction
KW - Least squares support vector machine
UR - http://www.scopus.com/inward/record.url?scp=85033460580&partnerID=8YFLogxK
U2 - 10.1145/3127404.3127449
DO - 10.1145/3127404.3127449
M3 - Conference contribution
AN - SCOPUS:85033460580
T3 - ACM International Conference Proceeding Series
SP - 225
EP - 228
BT - Proceedings - 12th Chinese Conference on Computer Supported Cooperative Work and Social Computing, ChineseCSCW 2017
PB - Association for Computing Machinery
T2 - 12th Chinese Conference on Computer Supported Cooperative Work and Social Computing, ChineseCSCW 2017
Y2 - 22 September 2017 through 23 September 2017
ER -
