Occupancy-Aware Spatio-Temporal Building Energy Forecasting with a Hybrid Long ShortTerm Memory and Graph Neural Network Benchmark Using Public Datasets

Open Access

Issue		E3S Web Conf. Volume 687, 2026 The 2^nd International Conference on Applied Sciences and Smart Technologies (InCASST 2025)


Article Number		02006
Number of page(s)		10
Section		Green Technologies & Digital Society
DOI		https://doi.org/10.1051/e3sconf/202668702006
Published online		15 January 2026

E. Delzendeh, S. Wu, A. Lee, and Y. Zhou, The impact of occupants’ behaviours on building energy analysis: A research review. Renew. Sustain. Energy Rev. 80, 1061–1071 (2017). https://doi.org/10.1016/j.rser.2017.05.264. [Google Scholar]
L. M. Candanedo and V. Feldheim, Accurate occupancy detection of an office room from light, temperature, humidity and CO₂ measurements using statistical learning models. Energy Build. 112, 28–39 (2016). https://doi.org/10.1016/j.enbuild.2015.11.071. [Google Scholar]
C. Beckel, W. Kleiminger, R. Cicchetti, T. Staake, and S. Santini, The ECO data set and the performance of non-intrusive load monitoring, in Proc. ACM BuildSys, 2014. https://doi.org/10.1145/2674061.2674064. [Google Scholar]
Y. Hu, X. Cheng, S. Wang, J. Chen, T. Zhao, and E. Dai, Times series forecasting for urban building energy consumption based on graph convolutional network. Appl. Energy. 307, 118231 (2022). https://doi.org/10.1016/j.apenergy.2021.118231. [Google Scholar]
L. Zhang et al., A review of machine learning in building load prediction, Appl. Energy. 285, 116452 (2021). https://doi.org/10.1016/j.apenergy.2021.116452. [Google Scholar]
J. Lee et al., Forecasting building operation dynamics using a physics-informed deep spatio-temporal graph neural network ensemble. Energy Build. 328, 115085 (2025). https://doi.org/10.1016/j.enbuild.2024.115085. [Google Scholar]
G. Li, Z. Yao, L. Chen, T. Li, and C. Xu, An interpretable graph convolutional neural network based method for fault diagnosis in building energy systems, Build. Simul. 17, 1113–1136 (2024). https://doi.org/10.1007/s12273-024-1125-6. [Google Scholar]
C. Miller et al., The Building Data Genome Project 2: Hourly energy meter data from the ASHRAE Great Energy Predictor III competition. Sci. Data. 7, 368 (2020). https://doi.org/10.1038/s41597-020-00712-x. [Google Scholar]
C. Miller et al., The ASHRAE Great Energy Predictor III competition: Overview and results. Sci. Technol. Built Environ. 26, 1427–1447 (2020). https://doi.org/10.1080/23744731.2020.1795514. [Google Scholar]
W. Kleiminger et al., ECO Data Set (Electricity Consumption & Occupancy), ETH Zürich Research Collection, 2016. https://www.research-collection.ethz.ch/entities/researchdata/e00ee826-4b2d-450c-8539-dc5ce14abc67. [Google Scholar]
ASHRAE Guideline 14-2014, Measurement of Energy, Demand, and Water Savings. American Society of Heating, Refrigerating and Air-Conditioning Engineers, 2014. [Google Scholar]
A. Nouri, A. Katramiz, P. Strachan, and H. GhaffarianHoseini, Statistical methodologies for verification of building energy models: Implications for Guideline-14 metrics, in Proc. IBPSA Building Simulation 2021, 2340–2347 (2021). https://doi.org/10.26868/25222708.2021.30538. [Google Scholar]
S. H. Ryu and H. J. Moon, Development of an occupancy prediction model using indoor environmental data based on machine learning techniques. Build. Environ. 107, 1–9 (2016). https://doi.org/10.1016/j.buildenv.2016.06.039. [Google Scholar]
Y. Ding, B. Zhou, Y. Chen, and X. Ouyang, Review on occupancy detection and prediction in building. Build. Simul. 15, 151–177 (2022). https://doi.org/10.1007/s12273-021-0813-8. [Google Scholar]
H. P. Das, S. Pal, and A. Bera, Machine learning for smart and energy-efficient buildings. Environ. Data Sci. 3, e44 (2024). https://doi.org/10.1017/eds.2024.34. [Google Scholar]

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.