- home
- about
- founder & publisher
- editorial boards
- advisory board
- for authors
- call for papers
- archive
- online first
- editorial policy
- participation fee
- contacts
THERMAL SCIENCE
International Scientific Journal
Find this paper on
VARIOUS ENSEMBLE METHODOLOGIES FOR BUILDING COOLING ENERGY USE PREDICTION
ABSTRACT
Due to the significant increase in cooling demand, accurately predicting cooling energy consumption in buildings has become essential. Machine learning techniques have been widely used for this purpose, with ensemble methods, which combine the outputs of multiple individual models, showing strong performance in handling its complexity. The primary objective of this study is to provide a systematic analysis of various ensemble aggregation strategies, including some novel combinations. Both homogeneous and heterogeneous ensembles are analyzed and tested using the same real-world measured dataset. Random forest and gradient boosting machine, which represent homogeneous ensembles that utilize bagging and boosting techniques, respectively, have proven to outperform individual models. Four prominent base models, namely feedforward neural network, decision tree, support vector regression, and gaussian process regression, are utilized to construct eight new heterogeneous ensembles. These base models are combined via weighted averaging, with various strategies for assigning weights systematically tested. The stacking/blending approach is also evaluated using all four base models along with multiple linear regression as meta learners. Results show that weighted averaging enhances prediction accuracy, with the best results obtained by solving an optimization problem that minimizes the mean squared error. Additional improvements are seen when an machine learning model acts as the aggregator, with the highest performance achieved using an feedforward neural network as the meta-learner. Despite its simplicity and low computational demands, multiple linear regression proves to be an effective meta-model, providing a solution that requires less extensive hyperparameter tuning compared to other nonlinear machine learning models.
KEYWORDS
PAPER SUBMITTED: 2025-08-05
PAPER REVISED: 2025-09-10
PAPER ACCEPTED: 2025-09-14
PUBLISHED ONLINE: 2025-11-01
DOI REFERENCE: https://doi.org/10.2298/TSCI250805178J
CITATION EXPORT: view in browser or download as text file
REFERENCES
[1] ***, Heat Roadmap Europe, Heating and Cooling Facts and Figures, https://publica-rest.fraunho-fer.de/server/api/core/bitstreams/969a5b7b-52bf-4e56-8fa6-97a13885e96b/content, 2017
[2] Santamouris, M., Cooling the Buildings-Past, Present and Future, Energy and Buildings, 128 (2016), Sept., pp. 617-638
[3] Zhang, L, et al., A Review of Machine Learning in Building Load Prediction, Applied Energy, 285 (2021), 116452
[4] Yang, Y., et al., A Survey on Ensemble Learning Under the Era of Deep Learning, Artificial Intelligence Review, 56 (2023), 6, pp. 5545-5589
[5] Wang, Z., et al., Identifying the Optimal Heterogeneous Ensemble Learning Model for Building Energy Prediction Using the Exhaustive Search Method, Energy and Buildings, 281 (2023), 112763
[6] Wu, T., et al., Evaluation of Stacking and Blending Ensemble Learning Methods for Estimating Daily Reference Evapotranspiration, Computers and Electronics in Agriculture, 184 (2021), 106039
[7] Yao, J., et al., Applications of Stacking/Blending Ensemble Learning Approaches for Evaluating Flash Flood Susceptibility, Int. Journal of Applied Earth Observation and Geoinformation, 112 (2022), 102932
[8] Jovanović, R. Ž., et al., Ensemble of Various Neural Networks for Prediction of Heating Energy Consumption, Energy and Buildings, 94 (2015), May, pp. 189-199
[9] Mohan, R., Pachauri, N., An Ensemble Model for the Energy Consumption Prediction of Residential Buildings, Energy, 314 (2025), 134255
[10] Jovanović, R., Sretenović, A., Various Multistage Ensembles for Prediction of Heating Energy Consumption, Modeling Identification and Control, 36 (2015), 2, pp. 119-132
[11] Jovanović, R. Ž., et al., Multistage Ensemble of Feedforward Neural Networks for Prediction of Heating Energy Consumption, Thermal Science, 20 (2016), 4, pp. 1321-1331
[12] Jovanović, R. Ž., Sretenović, A. A., Ensemble of Radial Basis Neural Networks with K-Means Clustering for Heating Energy Consumption Prediction, FME Transactions, 45 (2017), 1, pp. 1-7
[13] Cakiroglu, C., et al., Cooling Load Prediction of a Double-Story Terrace House Using Ensemble Learning Techniques and Genetic Programming with SHAP Approach, Energy and Buildings, 313 (2024), 114254
[14] Wang, Z., et al., A Comprehensive Review and Future Research Directions of Ensemble Learning Models for Predicting Building Energy Consumption, Energy and Buildings, 335 (2025), 115589
[15] Kezrane, C., et al., Artificial Neural Network Based Prediction of Engine-out Responses from a Biodiesel Fuelled Compression Ignition Engine, Thermal Science, 27 (2023), 4B, pp. 3433-3443
[16] Ji, S.-W., et al., Machine Learning-Based Evaluation of Acidizing Effectiveness and Optimization of Acidizing Parameters for Carbonate Gas Reservoir Horizontal Wells, Thermal Science, 29 (2025), 2A, pp. 1043-1048
[17] Lu, C., et al., Building Energy Prediction Using Artificial Neural Networks: A Literature Survey, Energy and Buildings, 262 (2022), 111718
[18] Wang, L., Support Vector Machines: Theory and Applications, vol. 177, Springer Science & Business Media, New York, USA, 2005
[19] Sretenović, A., et al., Support Vector Machine for the Prediction of Heating Energy Use, Thermal Science, 22 (2018), Suppl. 4, pp. S1171-S1181
[20] Rasmussen, C. E., Williams, C., Gaussian Processes for Machine Learning, The Mit Press, Cambridge, Mass., USA, 2006, vol. 32, pp. 68
[21] Alharbi, R., Algarni, A. D., Advanced Machine Learning Techniques for Predicting NOx Levels, Thermal Science, 28 (2024), 6B, pp. 4979-4989
[22] Bhinge, R., et al., Toward a Generalized Energy Prediction Model for Machine Tools, Journal of Manufacturing Science and Engineering, 139 (2017), 4, 041013
[23] Yoon, Y. R., Moon, H. J., Energy Consumption Model with Energy Use Factors of Tenants in Commercial Buildings using Gaussian Process Regression, Energy and Buildings, 168 (2018), June, pp. 215-224
[24] Yu, Z., et al., A Decision Tree Method for Building Energy Demand Modeling, Energy and Buildings, 42 (2010), 10, pp. 1637-1646
[25] Kang, W. H., et al., In-Situ Application of an ANN Algorithm for Optimized Chilled and Condenser Water Temperatures Set-Point During Cooling Operation, Energy and Buildings, 233 (2021), 110666
[26] Cao, L., et al., Electrical Load Prediction of Healthcare Buildings Through Single and Ensemble Learning, Energy Reports, 6 (2020), Nov., pp. 2751-2767
[27] Goyal, M., Pandey, M., A Systematic Analysis for Energy Performance Predictions in Residential Buildings Using Ensemble Learning, Arabian Journal for Science and Engineering, 46 (2021), 4, pp. 3155-3168
[28] Breiman, L., Random Forests, Machine Learning, 45 (2011), Oct., pp. 5-32
[29] Kor, H., Global Solar Radiation Prediction Model with Random Forest Algorithm, Thermal Science, 25 (2021), Special Issue 1, pp. 31-39
[30] Friedman, J. H., Greedy Function Approximation: A Gradient Boosting Machine, Annals of Statistics, 29 (2001), 5, pp. 1189-1232
[31] Shahhosseini, M., et al., Optimizing Ensemble Weights and Hyperparameters of Machine Learning Models for Regression Problems, Machine Learning with Applications, 7 (2022), 100251
[32] Perrone, M. P., Cooper, L. N., When Networks Disagree: Ensemble Methods for Hybrid Neural networks, in: How we Learn; How we Remember: Toward an Understanding of Brain and Neural Systems, World Scientific Series in 20th Century Physics, pp. 342-358, 1995
[33] Krogh, A., Vedelsby, J., Neural Network Ensembles, Cross Validation, and Active Learning, in: Advances in Neural Information Processing Systems, MIT Press, Cambridge, Mass., USA, Vol. 7, pp. 231-238, 1994
[34] Zhou, Z. H., et al., Ensembling Neural Networks: Many Could be Better Than All, Artificial Intelligence, 137 (2002), 1-2, pp. 239-263
[35] Shahhosseini, M., et al., Optimizing Ensemble Weights for Machine Learning Models: A Case Study for Housing Price Prediction, in: INFORMS International Conference on Service Science, Springer, New York, USA, 2019, pp. 87-97
[36] Boyd, S. P., Vandenberghe, L., Convex Optimization, Cambridge University Press, Cambridge, UK, 2004
[37] Opitz, D., Shavlik, J., Generating Accurate and Diverse Members of a Neural-Network Ensemble, in: Advances in Neural Information Processing Systems, MIT Press, Cambridge, Mass., USA, Vol. 8, pp. 535-543, 1995
[38] Sun, X., et al., Using the Inverse of Expected Error Variance to Determine Weights of Individual Ensemble Members: Application to Temperature Prediction, Journal of Meteorological Research, 31 (2017), 3, pp. 502-513
[39] Acar, E., Rais-Rohani, M., Ensemble of Metamodels with Optimized Weight Factors, Structural and Multidisciplinary Optimization, 37 (2009), Feb., pp. 279-294
[40] Breiman, L., Stacked Regressions, Machine Learning, 24 (1996), July, pp. 49-64
PDF VERSION [DOWNLOAD]
© 2026 Society of Thermal Engineers of Serbia. Published by the Vinča Institute of Nuclear Sciences, National Institute of the Republic of Serbia, Belgrade, Serbia. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International licence