Multi-Class EEG Classification for Brain Computer Interface based on CSP

The algorithm is presented, which is an extension of the Common Spatial Patterns(CSP) to multi-class case, in order to implement a suitable Brain-Computer Interface(BCI) system. An extension of the CSP algorithm to multiclass paradigms is introduced with an elaborate description. We use computer simulation to prove the algorithm right. After a short description of this algorithm, we will apply it in data set IlIa of BCI Competition III to prove the algorithm feasibility. The result illustrates the performance and validity of the algorithm. It not only improves the information transfer rate, but also assures the certain accuracy of the system.