Currently, the use of solar energy is increasing due to the rising demand for sustainable energy sources. Exploring solar energy presents several challenges, such as the unpredictability of weather conditions. Because of this, predicting solar energy production in advance is crucial. The integration of artificial intelligence into the field of renewable energy offers a solution to this challenge. However, this integration also has a challenge in selecting the appropriate algorithms to forecast solar energy production accurately. The focus of this study is the application of machine learning algorithms, specifically linear regression algorithms including Linear regression, Lasso regression, Ridge regression, and Elastic Net regression, combined with feature selection methods to forecast global horizontal irradiance. The performance metrics evaluated include the R2 score, Mean Absolute Error (MAE), and Root Mean Squared Error (RMSE). The dataset used was generated from a numerical simulation at latitude 22.55°N and longitude −14.33°E°. The results indicate that linear regression algorithms, when applied with various feature selection techniques, were not satisfactory due to the non-linear relationships among the weather variables. In the testing set, the R2 score was 0.29 with the correlation coefficient selection technique and 0.38 with L1 regularization, L2 regularization, and Elastic Net feature selection techniques. This study identifies gaps and provides opportunities for future research initiatives.