Performance Comparison of Random Forest Regressor and Support Vector Regression for Solar Energy Prediction

Abstract

Abstract This study conducts a comparative analysis between the Random Forest Regressor (RFR) and Support Vector Regression (SVR) for solar energy prediction. Solar energy, a prominent renewable energy source, but its predictability is challenging due to changing weather conditions. Micro grid operators, responsible for managing smaller, localized energy systems, often struggle to balance supply and demand efficiently because of the unpredictable nature of solar energy. To tackle this issue, we use machine learning models, RFR and SVR, with historical weather data including dew point, temperature, cloud cover, visibility and wind. Outcomes focus the superior performance of the Random Forest Regressor process compared to the Support Vector Regression in forecasting solar energy, demonstrating its potential for enhancing reliability in solar energy prediction models.