TY - GEN
T1 - EEG Classification with Broad Learning System and Composite Features
AU - Xu, Lincan
AU - Duan, Junwei
AU - Quan, Yujuan
AU - Zhou, Zhiguo
N1 - Publisher Copyright:
© 2021 IEEE.
PY - 2021/6/18
Y1 - 2021/6/18
N2 - Electroencephalogram (EEG) classification is one of the most important research topics of Brain Computer Interface (BCI). In this paper, a novel method based on broad learning system and composite features (CF-Bls) is proposed to deal with EEG data. Firstly, EEG signals are divided into 1-second 'frames' and mapped into 2D images. Then, Gabor filters are used to extract the texture features of the EEG images. After that, we extract abstract convolution features from the Gabor texture features and the EEG images by a convolutional neural network, respectively. Finally, the dimension of abstract convolution features is reduced by PCA and then the features are classified by broad learning system (BLS). Experimental results show that the accuracy and Kappa coefficient of CF-BLS have been significantly improved, compared with the existing algorithms.
AB - Electroencephalogram (EEG) classification is one of the most important research topics of Brain Computer Interface (BCI). In this paper, a novel method based on broad learning system and composite features (CF-Bls) is proposed to deal with EEG data. Firstly, EEG signals are divided into 1-second 'frames' and mapped into 2D images. Then, Gabor filters are used to extract the texture features of the EEG images. After that, we extract abstract convolution features from the Gabor texture features and the EEG images by a convolutional neural network, respectively. Finally, the dimension of abstract convolution features is reduced by PCA and then the features are classified by broad learning system (BLS). Experimental results show that the accuracy and Kappa coefficient of CF-BLS have been significantly improved, compared with the existing algorithms.
KW - Broad Learning System (BLS)
KW - Convolutional Neural Network (CNN)
KW - Gabor Filter
KW - Principal Component Analysis (PCA)
UR - http://www.scopus.com/inward/record.url?scp=85116450413&partnerID=8YFLogxK
U2 - 10.1109/SPAC53836.2021.9539966
DO - 10.1109/SPAC53836.2021.9539966
M3 - Conference contribution
AN - SCOPUS:85116450413
T3 - Conference Digest - 2021 International Conference on Security, Pattern Analysis, and Cybernetics, SPAC 2021
SP - 402
EP - 407
BT - Conference Digest - 2021 International Conference on Security, Pattern Analysis, and Cybernetics, SPAC 2021
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2021 International Conference on Security, Pattern Analysis, and Cybernetics, SPAC 2021
Y2 - 18 June 2021 through 20 June 2021
ER -