中国科学院机构知识库网格
Chinese Academy of Sciences Institutional Repositories Grid
首页 机构 成果 学者
  
  1. CAS IR Grid
  2. 自动化研究所
  3. 中国科学院自动化研究所
  4. 复杂系统认知与决策实验室
  5. 听觉模型与认知计算
EEGNet-based multi-source domain filter for BCI transfer learning

文献类型：期刊论文

作者Li, Mengfan1,2,3; Li, Jundi1,2,3; Song, Zhiyong1,2,3; Deng, Haodong1,2,3; Xu, Jiaming4,5; Xu, Guizhi1,2,3; Liao, Wenzhe6
刊名MEDICAL & BIOLOGICAL ENGINEERING & COMPUTING
出版日期2023-11-20
页码12
ISSN号0140-0118
关键词Brain-computer interface Multi-source domain filter Transfer learning Ensemble learning EEGNet
DOI10.1007/s11517-023-02967-z
通讯作者Li, Mengfan(mfli@hebut.edu.cn)
英文摘要Deep learning has great potential on decoding EEG in brain-computer interface. While common deep learning algorithms cannot directly train models with data from multiple individuals because of the inter-individual differences in EEG. Collecting enough data for each subject to satisfy the training of deep learning would result in an increase in training cost. This study proposes a novel transfer learning, EEGNet-based multi-source domain filter for transfer learning (EEGNet-MDFTL), to reduce the amount of training data and improve the performance of BCI. The EEGNet-MDFTL uses bagging ensemble learning to learn domain-invariant features from the multi-source domain and utilizes model loss value to filter the multi-source domain. Compared with baseline methods, the accuracy of the EEGNet-MDFTL reaches 91.96%, higher than two state-of-the-art methods, which demonstrates source domain filter can select similar source domains to improve the accuracy of the model, and remains a high level even when the data amount is reduced to 1/8, proving that ensemble learning learns enough domain invariant features from the multi-source domain to make the model insensitive to data amount. The proposed EEGNet-MDFTL is effective in improving the decoding performance with a small amount of data, which is helpful to save the BCI training cost.
WOS关键词NEURAL-NETWORK ; SYSTEM
资助项目Natural Science Foundation of Hebei Province[F2021202003] ; Technology Nova of Hebei University of Technology[JBKYXX2007] ; State Key Laboratory of Reliability and Intelligence of Electrical Equipment[EERI_OY2020004] ; State Key Laboratory of Reliability and Intelligence of Electrical Equipment[EERI_OY202000] ; National Natural Science Foundation of China[51977060] ; Key Research and Development Foundation of Hebei[19277752D] ; Key Research and Development Foundation of Hebei[21372002D]
WOS研究方向Computer Science ; Engineering ; Mathematical & Computational Biology ; Medical Informatics
语种英语
出版者SPRINGER HEIDELBERG
WOS记录号WOS:001104250600001
资助机构Natural Science Foundation of Hebei Province ; Technology Nova of Hebei University of Technology ; State Key Laboratory of Reliability and Intelligence of Electrical Equipment ; National Natural Science Foundation of China ; Key Research and Development Foundation of Hebei
源URL[http://ir.ia.ac.cn/handle/173211/54976]  
专题数字内容技术与服务研究中心_听觉模型与认知计算
通讯作者Li, Mengfan
作者单位1.Hebei Univ Technol, Sch Elect Engn, State Key Lab Reliabil & Intelligence Elect Equip, Tianjin 300401, Peoples R China
2.Hebei Key Lab Bioelectromagnet & Neuroengn, Tianjin 300132, Peoples R China
3.Hebei Univ Technol, Tianjin Key Lab Bioelect & Intelligent Hlth, Tianjin 300130, Peoples R China
4.Chinese Acad Sci, Inst Automat, Beijing, Peoples R China
5.Univ Chinese Acad Sci, Sch Artificial Intelligence, Beijing, Peoples R China
6.Hebei Univ Technol, Sch Artificial Intelligence, Tianjin, Peoples R China
推荐引用方式
GB/T 7714		 Li, Mengfan,Li, Jundi,Song, Zhiyong,et al. EEGNet-based multi-source domain filter for BCI transfer learning[J]. MEDICAL & BIOLOGICAL ENGINEERING & COMPUTING,2023:12.
APA Li, Mengfan.,Li, Jundi.,Song, Zhiyong.,Deng, Haodong.,Xu, Jiaming.,...&Liao, Wenzhe.(2023).EEGNet-based multi-source domain filter for BCI transfer learning.MEDICAL & BIOLOGICAL ENGINEERING & COMPUTING,12.
MLA Li, Mengfan,et al."EEGNet-based multi-source domain filter for BCI transfer learning".MEDICAL & BIOLOGICAL ENGINEERING & COMPUTING (2023):12.

入库方式： OAI收割

来源：自动化研究所

浏览0
下载0
收藏0
其他版本

除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。