Abstract
Today, electricity is in high demand in a variety of places, including hospitals, industry, households, transportation, and communication, among others. Renewable energy is a revolutionary type of energy that is increasingly being used to replace electricity demand because it has been regenerated and reused several times. Renewable energy is an intermediate and unpredictable natural resource, so it is difficult for many research studies to estimate its rate. To address this problem, this study uses a hybrid machine learning technique to precisely predict the energy level of natural resources. The hybrid machine learning is the combination of Multilayer Perceptron (MLP), Support Vector Regression (SVR) and CatBoost algorithm that increases the performance and predictability of renewable energy consumption. The proposed system dataset's results are evaluated at the train and test levels, and the results are then compared to other current approaches. The end results reveal that the proposed hybrid machine learning technique has a high prediction level when compared to others, as well as a lower cost rate and improved overall system performance.