Solar radiation prediction is essential for various applications, from renewable energy planning to weather fore-casting and environmental monitoring. Accurate solar radiation prediction can help improve the efficiency of energy production and facilitate better management of natural resources. This paper focuses on solar prediction as an essential parameter for weather forecasting. We applied several regression models to predict solar radiation. In conclusion, Random Forest Regression was the most accurate model for predicting solar radiation. The CNN model ranked third with an accuracy difference of approximately 0.16 compared to NN-MLP. However, there is a possibility of enhancing the accuracy of the CNN model by testing different combinations of layers and filters. Furthermore, the performance of the SVR model was improved by 1.577 after hyperparameter tuning. On the other hand, adjusting the number of estimators for the random forest regressor resulted in a slight decrease in performance, possibly due to over-fitting, as the tuning was done solely on the training data.
