Research Article
Time-Shift Correlation Algorithm for P300 Event Related Potential Brain-Computer Interface Implementation
Table 3
Performance comparison of the proposed approach and conventional CCA SVM.
| Methods | Classification accuracy/illegal trial rates (%) | Subject 1 | Subject 2 | Subject 3 | Proposed BCI | CCA SVM | Proposed BCI | CCA SVM | Proposed BCI | CCA SVM |
| Average 1 epoch | 90.0/30.0 | 87.92/40.0 | 84.81/21.0 | 83.52/46.0 | 75.00/28.0 | 74.00/50.0 | Data transfer rate (bit/min.) | 49.41 | 38.49 | 43.82 | 28.17 | 24.45 | 16.09 |
| Average 2 epochs | 95.35/14.0 | 96.22/15.50 | 88.64/12.0 | 91.44/18.50 | 83.72/14.0 | 87.1/19.0 | Data transfer rate (bit/min.) | 38.77 | 39.68 | 29.18 | 30.72 | 22.65 | 25.01 |
| Average 3 epochs | 100.0/3.03 | 100.0/7.0 | 96.67/9.09 | 97.78/9.25 | 90.0/9.09 | 90.48/12.0 | Data transfer rate (bit/min.) | 36.57 | 35.07 | 28.78 | 30.30 | 21.18 | 20.95 |
|
|
Accuracy rate = correct/(correct + incorrect). Illegal trial rate = illegal/total trials.
|