Research Article
Daily Crude Oil Price Forecasting Using Hybridizing Wavelet and Artificial Neural Network Model
Table 3
Training and testing performance of different network architecture of WANN model.
| Data | Network architecture | Training | Testing | Input | Hidden | RMSE | MAPE | RMSE | MAPE |
| Brent | 1 | 3 | 0.4437 | 1.46% | 0.8939 | 1.49% | 2 | 2 | 0.4225 | 1.35% | 0.8439 | 1.41% | 3 | 7 | 0.3756 | 1.24% | 0.8174 | 1.35% | 4 | 4 | 0.3736 | 1.20% | 0.8151 | 1.31% | 5 | 8 | 0.3489 | 1.15% | 0.8593 | 1.31% | 6 | 8 | 0.3431 | 1.14% | 0.9061 | 1.37% |
| WTI | 1 | 1 | 0.5108 | 1.60% | 0.8738 | 1.65% | 2 | 3 | 0.4617 | 1.46% | 0.8216 | 1.54% | 3 | 2 | 0.4252 | 1.37% | 0.7900 | 1.47% | 4 | 3 | 0.4171 | 1.33% | 0.7699 | 1.42% | 5 | 1 | 0.4044 | 1.29% | 0.7549 | 1.39% | 6 | 10 | 0.3712 | 1.24% | 0.9506 | 1.49% |
|
|