Table 4 Statistical results for soft computing in the estimations of cloudless DLR and the statistically significant models are in italics
From: Deep learning and regression modelling of cloudless downward longwave radiation
Model | Training phase (n = 143) | Testing phase (n = 62) | ||||||
|---|---|---|---|---|---|---|---|---|
r | RMSE | MBE | ts | r | RMSE | MBE | ts | |
NN | 0.8390 | 11.3324 | − 0.77E− 3 | 8.19E− 3 | 0.7556 | 12.3934 | 0.2565 | 0.1234 |
SVR-Gaussian | 0.8501 | 10.9932 | 1.6664 | 1.3802 | 0.7723 | 12.2118 | 1.9511 | 1.4567 |
SVR-linear | 0.8240 | 11.7666 | 1.3100 | 1.0083 | 0.7730 | 12.1311 | 1.2671 | 0.9452 |
SVR-polynomial (default) | 0.8327 | 11.6289 | 1.9928 | 1.5655 | 0.7727 | 12.3041 | 1.9035 | 1.4093 |
SVR-radial basic function | 0.8494 | 11.0089 | 1.6366 | 1.3530 | 0.7726 | 12.1926 | 1.8838 | 1.4074 |
ANFIS | 0.8359 | 11.3225 | 2.38E− 5 | 2.51E− 5 | 0.7932 | 11.5494 | 0.1825 | 0.1617 |