Published Papers

Volume 32 - Issue 3 (Theme Issue)
Theme Issue Papers

Machine learning models for window opening status and operation behaviours in sub‑tropical residential buildings

Tsz Wun Tsang, Bian Liu and Ling Tim Wong
Pages: 1-12Published: 30 Jul 2026
DOI: 10.33430/V32N3THIE‑2024‑0064
Cite thisHide

TSANG TW, LIU B, WONG LT, Machine learning models for window opening status and operation behaviours in sub‑tropical residential buildings, HKIE Transactions, Vol. 32, No. 3 (Theme Issue), Article THIE‑2024‑0064.R1, 2026, 10.33430/V32N3THIE‑2024‑0064

 Copy

Abstract:

Using a database of environmental observations on window status, window operation behaviours, and ambient environmental conditions, this study identified environmental drivers for window operation, offering robust predictive models for window opening status and window operation behaviour. Predictive models were developed using various algorithms, including logistic regression (LR), Gaussian Naive Bayes classifiers (GNBs), decision trees (DTs), random forests (RFs), and support vector machines (SVMs). It was found that SVMs, LR and GNBs demonstrated exemplary performance in predicting a simple task such as window opening status. However, these models struggled when faced with a multiclass classification problem indicating window operation behaviour. Conversely, tree‑based models effectively predicted window opening status and operation. Furthermore, by examining Gini importance (IG) scores, permutation importance (PIMP), and SHapley Additive exPlanation (SHAP) values, the study found that temperature and season were critical features in predicting window opening status, whereas the diurnal cycle exerted the strongest influence on window operation behaviour. Overall, this study systematically assessed the comparative performance of multiple machine learning approaches for predicting window status and operational behaviour, highlighting the importance of effective predictive model selection. The results contribute to advancing window behaviour modelling and improving the predictive reliability of building energy simulations.

Keywords:

window opening status; window operation behaviours; residential buildings;machine learning; environmental drivers

Reference List:

[1] Aloisio, A., Rosso, M.M., Di Battista, L., Quaranta, G., (2024). Machine‑learning‑aided regional postseismic
usability prediction of buildings꞉ 2016–2017 Central Italy earthquakes. Journal of Building Engineering, 91, 109526.
[2] Altmann, A., Toloşi, L., Sander, O., & Lengauer, T. (2010). Permutation importance꞉ a corrected feature importance measure. Bioinformatics, 26(10), 1340‑1347.
[3] Berrar, D. (2018). Bayes' theorem and naive Bayes classifier. Encyclopedia of bioinformatics and computational biology꞉ ABC of bioinformatics, 403, 412.
[4] Brager, G. S., & de Dear, R. J. (1998). Thermal adaptation in the built environment꞉ a literature review. Energy and Buildings, 27(1), 83‑96. https꞉//doi.org/Doi 10.1016/S0378‑7788(97)00053‑4
[5] Breiman, L. (2001). Random forests. Machine learning, 45, 5‑32.
[6] Chauhan, V. K., Dahiya, K., & Sharma, A. (2019). Problem formulations and solvers in linear SVM꞉ a review. Artificial Intelligence Review, 52(2), 803‑855.
[7] Chawla, N. V., Bowyer, K. W., Hall, L. O., & Kegelmeyer, W. P. (2002). SMOTE꞉ synthetic minority over‑sampling technique. Journal of artificial intelligence research, 16, 321‑357.
[8] Cortes, C., & Vapnik, V. (1995). Support‑vector networks. Machine learning, 20, 273‑297.
[9] Furuhashi, K., Nakaya, T., & Maeda, Y. (2022). Prediction of Occupant Behavior toward Natural Ventilation in Japanese Dwellings꞉ Machine Learning Models and Feature Selection. Energies, 15(16).
[10] Han, M. J., May, R., Zhang, X. X., Wang, X. R., Pan, S., Yan, D., & Jin, Y. (2020). A novel reinforcement learning method for improving occupant comfort via window opening and closing. Sustainable Cities and Society, 61.
[11] Hawila, A. A. W., Diallo, T. M., & Collignan, B. (2023). Occupants' window opening behavior in office building꞉ A review of influencing factors, modeling approaches and verification of literature models. Building and Environment, 110525.
[12] He, H., & Garcia, E. A. (2009). Learning from imbalanced data. IEEE Transactions on knowledge and data engineering, 21(9), 1263‑1284.
[13] Hosmer Jr, D. W., Lemeshow, S., & Sturdivant, R. X. (2013). Applied logistic regression (Vol. 398). John Wiley & Sons.
[14] Jeong, B., Jeong, J. W., & Park, J. S. (2016). Occupant behavior regarding the manual control of windows in residential buildings. Energy and buildings, 127, 206‑216.
[15] Jia, M., Srinivasan, R. S., & Raheem, A. A. (2017). From occupancy to occupant behavior꞉ An analytical survey of data acquisition technologies, modeling methodologies and simulation coupling mechanisms for building energy efficiency. Renewable and Sustainable Energy Reviews, 68, 525‑540.
[16] Jones, R. V., Fuertes, A., Gregori, E., & Giretti, A. (2017). Stochastic behavioural models of occupants' main bedroom window operation for UK residential buildings. Building and Environment, 118, 144‑158.
[17] Kc, R., Rijal, H. B., Shukuya, M., & Yoshida, K. (2018). An in‑situ study on occupants' behaviors for adaptive thermal comfort in a Japanese HEMS condominium. Journal of Building Engineering, 19, 402‑411.
[18] Kim, J., Hong, T., Jeong, J., Koo, C., & Kong, M. (2017). An integrated psychological response score of the occupants based on their activities and the indoor environmental quality condition changes. Building and Environment, 123, 66‑77.
[19] Lundberg, S. M., & Lee, S. I. (2017). A unified approach to interpreting model predictions. Advances in neural information processing systems, 30.
[20] Luo, M. H., Wang, Z., Ke, K., Cao, B., Zhai, Y. C., & Zhou, X. (2018). Human metabolic rate and thermal comfort in buildings꞉ The problem and challenge. Building and Environment, 131, 44‑52.
[21] Markovic, R., Grintal, E., Wölki, D., Frisch, J., & van Treeck, C. (2018). Window opening model using deep learning methods. Building and Environment, 145, 319‑329.
[22] Menze, B. H., Kelm, B. M., Masuch, R., Himmelreich, U., Bachert, P., Petrich, W., & Hamprecht, F. A. (2009). A comparison of random forest and its Gini importance with standard chemometric methods for the feature selection and
classification of spectral data. BMC bioinformatics, 10, 1‑16.
[23] Mo, H., Sun, H. J., Liu, J. J., & Wei, S. (2019). Developing window behavior models for residential buildings using XGBoost algorithm. Energy and Buildings, 205.
[24] Myles, A. J., Feudale, R. N., Liu, Y., Woody, N. A., & Brown, S. D. (2004). An introduction to decision tree
modeling. Journal of Chemometrics꞉ A Journal of the Chemometrics Society, 18(6), 275‑285.
[25] Nicol, J. F., & Humphreys, M. A. (2002). Adaptive thermal comfort and sustainable thermal standards for buildings. Energy and Buildings, 34(6), 563‑572.
[26] Pan, S., Han, Y., Wei, S., Wei, Y., Xia, L., Xie, L., ... & Yu, W. (2019). A model based on Gauss Distribution for predicting window behavior in building. Building and Environment, 149, 210‑219.
[27] Park, J., Jeong, B., Chae, Y. T., & Jeong, J. W. (2021). Machine learning algorithms for predicting occupants' behaviour in the manual control of windows for cross‑ventilation in homes. Indoor and Built Environment, 30(8), 1106‑1123.
[28] Rijal, H. B. (2021). Thermal adaptation of buildings and people for energy saving in extreme cold climate of Nepal. Energy and Buildings, 230.
[29] Rijal, H. B., Humphreys, M. A., & Nicol, J. F. (2018). Development of a window opening algorithm based on adaptive thermal comfort to predict occupant behavior in Japanese dwellings. Japan Architectural Review, 1(3), 310‑321.
[30] Rijal, H. B., Tuohy, P., Humphreys, M. A., Nicol, J. F., Samuel, A., & Clarke, J. (2007). Using results from field surveys to predict the effect of open windows on thermal comfort and energy use in buildings. Energy a n d B u i l d i n g s , 3 9 ( 7 ) , 8 2 3 ‑ 8 3 6. https꞉//doi.org/10.1016/j.enbuild.2007.02.003
[31] Rouleau, J., & Gosselin, L. (2020). Probabilistic window opening model considering occupant behavior diversity꞉ A data‑driven case study of Canadian residential buildings. Energy, 195, 116981.
[32] Strobl, C., Boulesteix, A. L., & Zeileis, A. Hothorn, és T. (2007). Bias in random forest variable importance
measures꞉ illustrations, sources and a solution. BMC Bioinformatics, 8, 25.
[33] Tsang, T. W., Wong, L. T., Lung, H. T., & Mui, K. W. (2024). Stochastic behavioral models of bedroom window operation in sub‑tropical residential buildings. Building and Environment, 262, 111784.
[34] Wei, Y., Yu, H., Pan, S., Xia, L., Xie, J., Wang, X., ... & Li, Q. (2019). Comparison of different window behavior modeling approaches during transition season in Beijing, China. Building and Environment, 157, 1‑15.
[35] Yang, X., Liu, J., Meng, Q., Wei, Y., Lei, Y., Wu, M., ...& Lian, Y. (2022). Window Opening Behavior of Residential Buildings during the Transitional Season in China's Xi'an. Complexity, 2022.
[36] Yao, M., & Zhao, B. (2017). Window opening behavior of occupants in residential buildings in Beijing. Building and Environment, 124, 441‑449.
[37] Zaki, S. A., Damiati, S. A., Rijal, H. B., Hagishima, A., & Abd Razak, A. (2017). Adaptive thermal comfort in university classrooms in Malaysia and Japan. Building and Environment, 122, 294‑306.
[38] Zelaya, C. V. G. (2019, April). Towards explaining the effects of data preprocessing on machine learning. In 2019 IEEE 35th international conference on data engineering (ICDE) (pp. 2086‑2090). IEEE.
>> more<< less