Table 4. SVM with RBF kernel using 6 features.
c–value | Gamma γ | Training accuracy (%) | Testing accuracy (%) |
---|---|---|---|
0.5 | 0.1 | 85.89 | 81.16 |
0.7 | 87.13 | 81.15 | |
0.9 | 87.55 | 82.60 | |
1 | 87.96 | 82.60 | |
2 | 89.21 | 82.60 | |
5 | 89.62 | 81.15 | |
7 | 90.04 | 82.60 | |
0.9 | 0.1 | 87.55 | 82.60 |
0.2 | 88.38 | 81.15 | |
0.3 | 89.62 | 81.15 | |
0.5 | 90.87 | 78.26 | |
0.6 | 91.28 | 76.81 | |
0.9 | 93.36 | 76.81 |