Table 3 Comparison of relative humidity prediction model performance.
From: Evaluation of climate prediction models in Yunnan, China: traditional methods and AI approaches
RMSE | MAE | R | |||||||
|---|---|---|---|---|---|---|---|---|---|
Train | Validation | Test | Train | Validation | Test | Train | Validation | Test | |
RegCM | 4.7650 | 4.8543 | 4.8233 | 3.2809 | 3.0585 | 3.0280 | 0.8542 | 0.8387 | 0.8443 |
CNN | 4.5650 | 4.6281 | 4.6327 | 3.7951 | 3.8210 | 3.8308 | 0.9367 | 0.9230 | 0.9237 |
LSTM | 4.6546 | 4.8131 | 4.8353 | 3.6107 | 3.7388 | 3.7588 | 0.9483 | 0.9448 | 0.9459 |
Transformer | 7.7144 | 7.7876 | 7.8586 | 5.8971 | 5.9487 | 6.0399 | 0.8505 | 0.8476 | 0.8518 |
CNN-LSTM | 3.6847 | 3.8610 | 3.9061 | 2.8789 | 3.0313 | 3.0591 | 0.9679 | 0.9648 | 0.9649 |
LSTM-Transformer | 3.5414 | 3.6833 | 3.7054 | 2.7647 | 2.8739 | 2.8825 | 0.9728 | 0.9704 | 0.9710 |
