Table 7.
The effect of secondary feature selection on average classification accuracy of various datasets (the LOG method).
| Datasets | Without Secondary Feature Selection | With Secondary Feature Selection | ||||
|---|---|---|---|---|---|---|
| CSP-Wavelet+LOG | CSP-WPD+LOG | CSP-FB+LOG | CSP-Wavelet+LOG | CSP-WPD+LOG | CSP-FB+LOG | |
| dataset 1 | 80.71 | 81.57 | 81.71 | 84.29 | 85 | 88.86 |
| dataset 2 | 80.63 | 80.4 | 79.22 | 83.4 | 83.05 | 82.15 |
| dataset 3 | 75.97 | 76.11 | 79.17 | 78.19 | 78.75 | 81.53 |
| dataset 4 | 77.38 | 77.86 | 74.76 | 79.4 | 80.83 | 80.12 |
| All data | 79.6 | 79.62 | 78.68 | 82.25 | 82.38 | 82.3 |