A Tutorial on Support Vector Machines for Pattern Recognition
Data Mining and Knowledge Discovery
Hi-index | 0.00 |
In order to build a high-performance brain-computer interface (BCI) for cursor movement control, a P300-based BCI system using a five-select oddball paradigm was designed and implemented. We found that high intensity visual stimuli (HIVS) can improve the performance of BCI. 9 subjects participated in the test of the proposed BCI system. Each subject completed 40 epochs with HIVS and low intensity visual stimuli (LIVS) respectively. The preprocessed data were classified by support vector machines (SVM). The averaged waveforms both from HIVS and LIVS proved that this new paradigm can elicit evident P300 potentials. Furthermore, the results indicated the information transfer rate (ITR) of HIVS could reach 5.4 bit/min, which was higher than 4.6 bit/min of LIVS.