Table 6 The classification accuracy of the KNN with feature selection (VMD-based FE methods).
No. of Features | Classification accuracy % (no. of channels) | |||||
|---|---|---|---|---|---|---|
VMD + KFD | VMD + TeEng | VMD + SuEn | VMD + LBP | VMD + IQR | VMD + ShEn | |
1 | – | 61.94 | 65.42 | 55.21 | 57.22 | – |
2 | 73.75 | 87.01 | 81.39 | 80.35 | 80.97 | – |
3 | 81.67 | 88.68 | 84.10 | 84.79 | 84.72 | 82.92 |
4 | 87.57 | 91.32 | 88.26 | 86.39 | 87.78 | 84.58 |
5 | 89.31 | 91.94 | – | 88.33 | 88.13 | 85.55 |
6 | 90.83 | 92.22 | 88.89 | 88.96 | 89.24 | 85.83 |
7 | 91.73 | – | 90.21 | 90.42 | – | 86.67 |
8 | 91.74 | 92.29 | 90.97 | 90.90 | – | 88.06 |
9 | 92.22 | 92.43 | 91.11 | 91.11 | – | 88.40 |
10 | – | – | 91.53 (6) | 91.25 | 89.51 | 88.47 |
11 | – | 92.57 | – | 91.39 | – | 88.68 |
12 | – | 92.71 (6) | – | 91.53 | 89.65 | 88.96 |
13 | – | – | – | 91.67 | 89.86 (9) | - |
14 | – | – | – | – | – | 89.10 (8) |
15 | – | – | – | 91.74 | – | – |
16 | 92.29 | – | – | – | – | – |
17 | 92.36 | – | – | – | – | – |
18 | – | – | – | 91.88 | – | – |
19 | 92.57 (11) | – | – | – | – | – |
20 | – | – | – | 91.94 (11) | – | – |
All (133) | 53.26 (19) | 55.28 (19) | 62.36 (19) | 63.61 (19) | 67.15 (19) | 69.24 (19) |
