TY - JOUR
T1 - Linear and nonlinear methods for brain-computer interfaces
AU - Müller, Klaus Robert
AU - Anderson, Charles W.
AU - Birch, Gary E.
N1 - Funding Information:
Manuscript received September 4, 2002; revised May 24, 2003. The work of K.-R. Müller was supported in part by the Deutsche Forschungsgemeinschaft (DFG) under contracts JA 379/9-1 and JA 379/7-1 and in part by the Bundesmin-isterium fuer Bildung und Forschung (BMBF) under contract FKZ 01IBB02A. The work of C. Anderson was supported in part by the National Science Foundation under Grant 9202100. The work of G. Birch was supported in part by the National Sciences and Engineering Research Council of Canada (NSERC) under Grant 90278-2002.
PY - 2003/6
Y1 - 2003/6
N2 - At the recent Second International Meeting on Brain-Computer Interfaces (BCIs) held in June 2002 in Rensselaerville, NY, a formal debate was held on the pros and cons of linear and nonlinear methods in BCI research. Specific examples applying EEG data sets to linear and nonlinear methods are given and an overview of the various pros and cons of each approach is summarized. Overall, it was agreed that simplicity is generally best and, therefore, the use of linear methods is recommended wherever possible. It was also agreed that nonlinear methods in some applications can provide better results, particularly with complex and/or other very large data sets.
AB - At the recent Second International Meeting on Brain-Computer Interfaces (BCIs) held in June 2002 in Rensselaerville, NY, a formal debate was held on the pros and cons of linear and nonlinear methods in BCI research. Specific examples applying EEG data sets to linear and nonlinear methods are given and an overview of the various pros and cons of each approach is summarized. Overall, it was agreed that simplicity is generally best and, therefore, the use of linear methods is recommended wherever possible. It was also agreed that nonlinear methods in some applications can provide better results, particularly with complex and/or other very large data sets.
KW - Feature spaces
KW - Fisher's discriminant
KW - Linear methods
KW - Mathematical programming machines
KW - Support vector machines (SVMs)
UR - http://www.scopus.com/inward/record.url?scp=0041353047&partnerID=8YFLogxK
U2 - 10.1109/TNSRE.2003.814484
DO - 10.1109/TNSRE.2003.814484
M3 - Article
C2 - 12899264
AN - SCOPUS:0041353047
SN - 1534-4320
VL - 11
SP - 165
EP - 169
JO - IEEE Transactions on Neural Systems and Rehabilitation Engineering
JF - IEEE Transactions on Neural Systems and Rehabilitation Engineering
IS - 2
ER -