Analysis and classification of EEG signals using a hybrid clustering technique
Lecturer:	Dr. Yan Li Department of Mathematics and Computing, University of Southern Queensland, Australia
Date:	January 7, 2011(Friday) 13:30-14:30
Room:	Okayama University
Summary: This talk presents a new classification approach using clustering technique-based least square support vector machine for the classification of EEG signals. Decision making is performed in two stages. Firstly, clustering technique is employed to extract representative features of EEG data. Secondly, least square support vector machine is applied to the extracted features to classify two-class EEG signals. Extensive experiments are conducted on benchmark EEG databases to verify the effectiveness of the proposed method. Comparisons of the approach with other existing methods for the same databases are also carried out. The experimental results demonstrate that the proposed algorithm produces a better classification rate than the previous reported methods and takes much less computing time. The research findings in this study have been accepted for publishing in top journals.