Table 4.
Error analysis results of different multi-kernel support vector regression models
| Kernel function | RMSE | MAE | MAPE (%) |
|---|---|---|---|
| Linear | 93.88 | 79.59 | 0.84 |
| RBF | 878.03 | 859.19 | 8.95 |
| Poly | 751.88 | 728.42 | 7.58 |
| Sigmoid | 1697.99 | 1692.36 | 17.68 |
| EGMPA-Linear | 93.88 | 79.59 | 0.84 |
| EGMPA-RBF | 135.64 | 121.22 | 1.28 |
| EGMPA-Poly | 97.98 | 79.22 | 0.83 |
| EGMPA-Sigmoid | 130.91 | 115.84 | 1.22 |
| EGMPA-Linear-Poly | 89.92 | 74.48 | 0.78 |
| EGMPA-Linear-RBF | 93.88 | 79.59 | 0.84 |
| EGMPA-Linear-Sigmoid | 87.78 | 72.26 | 0.76 |
| EGMPA-Poly-RBF | 57.27 | 46.75 | 0.49 |
| EGMPA-Poly-Sigmoid | 89.95 | 74.48 | 0.78 |
| EGMPA-RBF-Sigmoid | 37.43 | 30.63 | 0.32 |
Note: Bold values in the table represent the best values