Table 6 The generalization ability of two machine learning models
From: Experimental study on creep properties prediction of reed bales based on SVR and MLP
Applied Forces (kN) | Model | MAE | RMSE | R | R2 |
|---|---|---|---|---|---|
120 | MLP | 2.110 × 10–3 | 2.692 × 10–3 | 0.9992 | 0.9989 |
SVR | 1.417 × 10–4 | 1.612 × 10–4 | 0.9941 | 0.7730 | |
160 | MLP | 9.878 × 10–4 | 1.354 × 10–3 | 0.9354 | 0.9115 |
SVR | 1.486 × 10–4 | 1.516 × 10–4 | 0.9939 | 0.6416 | |
200 | MLP | 1.962 × 10–3 | 2.539 × 10–3 | 0.9989 | 0.9979 |
SVR | 1.294 × 10–4 | 1.374 × 10–4 | 0.9851 | 0.7058 | |
240 | MLP | 2.158 × 10–3 | 2.675 × 10–3 | 0.9993 | 0.9985 |
SVR | 1.588 × 10–4 | 1.839 × 10–4 | 0.9892 | 0.8921 | |
280 | MLP | 9.796 × 10–4 | 1.157 × 10–3 | 0.9252 | 0.9047 |
SVR | 4.144 × 10–4 | 4.378 × 10–4 | 0.9885 | 0.8759 |