Table 5 Classification accuracy and execution time of the proposed feature set with a varied number of VMFs using dataset I.
No. of VMFs | Dim. N = no. of segments (rows) | Execution Time(s) (2000 Hz sampling 25 s) | Accuracy (%) | |||
|---|---|---|---|---|---|---|
SVM | KNN | Decision Tree | Random Forest | |||
2 | Nx6 | 0.31 ± 0.01 | 71.12 ± 2.09 | 89.12 ± 0.78 | 84.93 ± 2.85 | 90.34 ± 0.72 |
4 | Nx12 | 0.37 ± 0.013 | 79.86 ± 0.76 | 93.71 ± 0.50 | 89.12 ± 0.48 | 93.64 ± 0.49 |
6 | Nx18 | 0.50 ± 0.03 | 82.37 ± 0.71 | 95.32 ± 0.58 | 89.29 ± 0.51 | 95.13 ± 0.31 |
8 | Nx24 | 0.77 ± 0.04 | 83.64 ± 0.93 | 95.74 ± 0.42 | 89.76 ± 0.81 | 94.97 ± 0.35 |
10 | Nx30 | 0.91 ± 0.04 | 85.80 ± 0.61 | 96.92 ± 0.50 | 91.84 ± 0.48 | 95.51 ± 0.43 |
12 | Nx36 | 1.03 ± 0.06 | 86.8 ± 1.09 | 97.45 ± 0.33 | 93.37 ± 0.52 | 97.00 ± 0.21 |
