A Lightweight Multi-Layer Perceptron Approach for Carbon Emission Prediction of Public Buildings Under Low-Dimensional Data Scenarios
Open Access
- 12 December 2025
- Vol. 15 (24) , 4508
Abstract
Amid global efforts toward carbon neutrality, carbon emission accounting in the construction sector has become essential for sustainable design. Public buildings, with complex energy systems and high operational loads, are major carbon emitters. However, early design stages often provide only low-dimensional parameters—such as floor area, number of floors, and location—limiting conventional regression methods. This study develops a lightweight prediction framework using a multilayer perceptron (MLP) neural network. Feature engineering constructs composite indicators—layers per unit area (LPA) and height-to-area ratio (HAR)—to quantify spatial compactness and vertical density. A three-layer MLP with Swish activation, adaptive L2 regularization, and Dropout reduces overfitting and improves generalization. Tests show the model achieves a mean absolute error of 4160 tCO2 and R2 of 0.966, reducing prediction error by 54.7% compared to linear regression. For high-rise buildings (>15 floors), error remains below 8.1%. SHAP analysis highlights floor area as the dominant factor (51.2%), while HAR and LPA jointly improve accuracy by 5.8%. A Python-based tool is developed for rapid emission estimation during design. Using 150 samples and 10-fold cross-validation, this work demonstrates the potential of deep learning in low-dimensional carbon prediction, offering a practical reference for early-stage green building design, though generalizability requires further validation with larger datasets.Keywords
This publication has 12 references indexed in Scilit:
- Structural translational-torsional coupling response control using Transformer-based bidirectional semi-active TMDJournal of Building Engineering, 2025
- Semi-active variable stiffness and damping control for adjacent structures using LSTM-based prediction algorithmJournal of Building Engineering, 2025
- Spatio-temporal distribution and peak prediction of energy consumption and carbon emissions of residential buildings in ChinaApplied Energy, 2024
- Utilizing weather forecast meteorological models for building energy simulations: A case study of a multi-unit residential complexEnergy and Buildings, 2023
- Seismic control of adaptive variable stiffness intelligent structures using fuzzy control strategy combined with LSTMJournal of Building Engineering, 2023
- What is the driving mechanism for the carbon emissions in the building sector? An integrated DEMATEL-ISM modelEnergy, 2023
- Carbon emissions in China's urban residential building sector through 2060: A dynamic scenario simulationEnergy, 2022
- A simple approach for short-term wind speed interval prediction based on independently recurrent neural networks and error probability distributionEnergy, 2022
- CO 2 emissions in China's building sector through 2050: A scenario analysis based on a bottom-up modelEnergy, 2017
- A review of recent multi-region input–output models used for consumption-based emission and resource accountingEcological Economics, 2009