Table 1 Prediction accuracy comparison with different kernel functions and parameters. The results were obtained by 5-fold cross-validation.
SVM model | Kernel function | Parameters | Accuracy (%) |
|---|---|---|---|
1 | Polynomial | α = 1, β = 1, d = 2 | 59.0 |
2 | Polynomial | α = 1, β = 1, d = 5 | 60.5 |
3 | RBF | γ = 0.01, C = 2.0 | 62.8 |
4 | RBF | γ = 0.06, C = 2.0 | 60.5 |
5 | RBF | γ = 0.2, C = 1.0 | 62.6 |