Ensemble Machine Learning Models Improve Predictions for Residential Energy Consumption

Researchers from the Department of Computer Science have made significant contributions to the field of applied computational intelligence and soft computing. Their study presents four ensemble machine learning models for predicting residential energy consumption in South Africa, leveraging historical energy consumption patterns and enhancing predictive abilities through feature engineering methodologies.

Key Takeaways:

• The study presents four ensemble machine learning models: ensemble by averaging (EA), ensemble by stacking each estimator (ESE), ensemble by boosting (EB), and ensemble by voting estimator (EVE), built on top of Random Forest (RF) and Decision Tree (DT) base predictor models.
• The accuracy of each ensemble model was evaluated using performance indicators such as mean squared error (MSE), mean absolute error (MAE), mean absolute percentage error (MAPE), and coefficient of determination R2.
• The study provides valuable insights for researchers and practitioners in predicting energy consumption in residential buildings and highlights the benefits of using ensemble learning models in the building and energy research domains.
• The study demonstrated the efficiency of ensemble learning models in providing accurate predictions for residential energy consumption.
• The researchers used feature engineering methodologies to enhance the predictive abilities of the base predictor models, leveraging historical energy consumption patterns to capture temporal intricacies.
• The department of Computer Science, led by David Attipoe, supported this research with financial assistance from the University of South Africa.

Statistics:

• The study presented four ensemble machine learning models for predicting residential energy consumption.
• The accuracy of each ensemble model was evaluated using performance indicators such as MSE, MAE, MAPE, and R2.
• The study demonstrated the efficiency of ensemble learning models in providing accurate predictions, with the coefficient of determination R2 indicating the proportion of the variance in the dependent variable that is predictable from the independent variable(s).

Sources:

• NewsRx. Department of Computer Science Researchers Report on Findings in Applied Computational Intelligence and Soft Computing (Predicting Residential Energy Consumption in South Africa Using Ensemble Models). Information Technology Newsweekly. May 20, 2025; p 144.
• Applied Computational Intelligence and Soft Computing. Predicting Residential Energy Consumption in South Africa Using Ensemble Models. (https://www.hindawi.com/journals/acisc/)
• Wiley. Applied Computational Intelligence and Soft Computing.