Table 3 Findings of the comparative analysis between the PBNN and other techniques.
From: Parallel boosting neural network with mutual information for day-ahead solar irradiance forecasting
Location | Model | RMSE | MAE | MSE | MAPE | NRMSE | RMSRE | MARE | RMSPE |
|---|---|---|---|---|---|---|---|---|---|
(\((W~m^{-2})\)) | (\((W~m^{-2})\)) | (\((W~m^{-2})^2\)) | (\((\%)\)) | (\((\%)\)) | (\((\%)\)) | ||||
Islamabad | CatBoost | 30.67 | 21.47 | 940.69 | 0.68 | 0.64 | 0.014 | 0.0068 | 1.41 |
XgBoost | 22.98 | 13.17 | 528.37 | 0.54 | 0.48 | 0.015 | 0.0054 | 1.55 | |
RF | 16.13 | 9.61 | 260.13 | 0.31 | 0.34 | 0.008 | 0.003 | 0.81 | |
LSTM | 38.57 | 28.59 | 1487.91 | 1.04 | 0.79 | 0.025 | 0.01 | 2.54 | |
GRU | 55.21 | 47.04 | 3048.05 | 1.42 | 1.15 | 0.022 | 0.014 | 2.18 | |
PBNN | 14.06 | 8.36 | 197.77 | 0.26 | 0.29 | 0.006 | 0.002 | 0.56 | |
San Diego | CatBoost | 38.68 | 29.69 | 1496.52 | 0.63 | 0.74 | 0.008 | 0.006 | 0.86 |
XgBoost | 19.19 | 10.51 | 368.22 | 0.23 | 0.37 | 0.004 | 0.002 | 0.48 | |
RF | 26.26 | 11 | 689.45 | 0.24 | 0.5 | 0.006 | 0.002 | 0.65 | |
LSTM | 75.38 | 47.41 | 5682.16 | 0.99 | 1.45 | 0.015 | 0.009 | 1.56 | |
GRU | 78.98 | 50.78 | 6239.16 | 1.08 | 1.51 | 0.017 | 0.011 | 1.73 | |
PBNN | 17.23 | 5.26 | 296.84 | 0.12 | 0.32 | 0.004 | 0.001 | 0.48 |
