Meteorological Variables Forecasting System Using Machine Learning and Open-Source Software

The techniques for forecasting meteorological variables are highly studied since prior knowledge of them allows the efficient management renewable energies, and also other applications science such as agriculture, health, engineering, energy, etc. In this research, design, implementation, comparison models have been performed using different Machine Learning part Python open-source software. implemented include multiple linear regression, polynomial random forest, decision tree, XGBoost, multilayer perceptron neural network (MLP). To identify best technique, mean square error (RMSE), absolute percentage (MAPE), (MAE), coefficient determination (R2) used evaluation metrics. most depend on variable to be forecasting, however, it is noted that them, forest XGBoost present better performance. For temperature, performing technique was Random Forest with an R2 0.8631, MAE 0.4728 °C, MAPE 2.73%, RMSE 0.6621 °C; relative humidity, 0.8583, 2.1380RH, 2.50% 2.9003 RH; solar radiation, 0.7333, 65.8105 W/m2, 105.9141 W/m2; wind speed, 0.3660, 0.1097 m/s, 0.2136 m/s.