Research on EEG based on SVM and EMD

Xinxin Wang, Jianlin Zhao

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

To provide accurate judgment of epilepsy for doctors, this paper made a study of two methods to classify epileptic electroencephalogram (EEG) according to the difference of the waveform and energy characteristics between EEG and normal EEG. One was adopting Support Vector Machines (SVM), the other was using the combination of the Empirical Mode Decomposition (EMD) and SVM, and the accuracy rate of epilepsy EEG and the static epilepsy EEG was compared. The experimental results indicate that the second method can achieve better effect on the classification of EEG, and distinguish effectively epileptic EEG and normal EEG. The innovation of this study is that it exerts the method of EEG based on SVM and EMD effectively.

Original languageEnglish
Title of host publicationInformation Computing and Applications - Third International Conference, ICICA 2012, Proceedings
Pages745-751
Number of pages7
EditionPART 2
DOIs
StatePublished - 2012
Externally publishedYes
Event3rd International Conference on Information Computing and Applications, ICICA 2012 - Chengde, China
Duration: 14 Sep 201216 Sep 2012

Publication series

NameCommunications in Computer and Information Science
NumberPART 2
Volume308 CCIS
ISSN (Print)1865-0929

Conference

Conference3rd International Conference on Information Computing and Applications, ICICA 2012
Country/TerritoryChina
CityChengde
Period14/09/1216/09/12

Keywords

  • Classification
  • EEG
  • EMD
  • SVM

Fingerprint

Dive into the research topics of 'Research on EEG based on SVM and EMD'. Together they form a unique fingerprint.

Cite this