Advancing Sustainability Through Machine Learning: Modeling and Forecasting Renewable Energy Consumption

This research provides a thorough examination of the industrial sector’s forecasting of renewable energy consumption, utilizing sophisticated machine learning techniques to enhance the accuracy and reliability of the predictions. LASSO regression, random forest (RF), Support Vector Regression (SVR), eXtreme Gradient Boosting (XGBoost 2.1.3), LightGBM, and multilayer perceptron (MLP) were all selected due to their ability to effectively handle large datasets. Our primary goal was to demonstrate the utility of the Energy Uncertainty Index (EUI) within commonly accepted models to ensure replicability and relevance to a broad audience. The integration of the EUI as an independent variable is a critical innovation of this research, as it addresses the challenges presented by fluctuations in energy markets. A more nuanced comprehension of consumption trends in the presence of uncertainty is achieved through this inclusion. We evaluate the performance of these models in the context of renewable energy consumption forecasting, identifying their strengths and limitations. The results indicate that the prognostic potential of the models is considerably improved by the inclusion of the EUI, providing valuable insights for energy policymakers, investors, and industry stakeholders. These advancements emphasize the role of machine learning in achieving efficient resource allocation, guiding infrastructure development, minimizing risks, and supporting the global transition toward renewable energy and sustainability.