隨著全球邁向2050淨零排放，建築全生命週期評估（WLCA）已成為各國減碳策略的核心焦點。 WLCA 涵蓋營運碳排（OC）與蘊含碳排（EC）兩個範疇。過去數十年的政策多側重於 OC 的管制；直到 2010 年代，隨著建築能源效率提升， EC 的重要性才逐漸受到重視，然而，相較於發展成熟的 OC，目前各國對 EC 的評估標準與研究仍顯匱乏。臺灣雖已於2024年公告實施低碳（低蘊含碳）建築標示制度（LEBR），並採用碳排減碳率（CFR）作為分級之評估指標，分為 1+ 級至 7 級共8個等級，但現行分級基準仍缺乏足夠的案例驗證，難以精準反映不同建築減碳的潛力。文獻指出，藉由各類型建築蘊含碳強度（ECI）的分析統計，不僅能用於分級基準分析，亦能用於預測開發行為的蘊含碳排，協助設計者於規劃階段即可導入減碳策略。
本研究旨在建立一套系統化的建築蘊含碳排基準分析流程，並開發適用於建築設計初期的機器學習預測工具。研究將以臺灣集合住宅作為實證案例，針對原始樣本有限之困境，導入拉丁超立方抽樣（LHS）進行資料增強，並配合統計檢定方法，確保模擬數據集的代表性。結果顯示，模擬數據之中位數（458.42 kgCO2e/m2）與原始案例（470.83 kgCO2e/m2）落差僅約 2.64 %，且透過 K-S 檢定等統計驗證，證實該流程能有效擴展數據規模，克服小樣本數據之限制。
統計分析顯示，臺灣集合住宅之ECI呈現右偏分佈，其模擬樣本中位數為 458.42 kgCO2e/m2，可作為產業評估的參考基準線（Baseline）。此外，本研究進一步討論CSER 減碳情境模擬，未來若低碳混凝土普及率提升至 80%，該基準線可下移 3.38%（中位數442.92 kgCO2e/m2）。在分級制度與政策建議上，本研究提出兩種ECI分級策略：獎勵型政策建議採用「累積百分比法」，以識別前 20% 的低碳建築案例；而強制型政策則建議採用「信賴區間法」，以產業平均值作為分級基準。
在預測工具的開發方面，本研究探討並整合多種機器學習演算法，建立一套在建築規劃初期快速評估蘊含碳強度（ECI）的評估架構。結果顯示，支持向量迴歸模型（SVR）在預測效能上最為穩定，其測試集解釋力（R-squared）達 0.760，且平均絕對百分比誤差（MAPE）控制在 3.5% 以內，證實了關鍵建築特徵與 ECI 之間存在顯著相關。該預測工具能讓設計者或開發者在建築規劃初期僅透過關鍵設計參數，就能精確掌握ECI 的分佈區間，為建築減碳提供科學化的決策支持。

As global efforts align with the 2050 net-zero target, Whole Life Carbon Assessment (WLCA) has become pivotal. However, policy and research on Embodied Carbon (EC) remain scarce compared to Operational Carbon (OC). Although Taiwan implemented the Low Embodied Carbon Building Rating (LEBR) system in 2024 using the Carbon Reduction Rate (CFR) as a metric, its baseline still lacks sufficient empirical validation. Literature suggests that statistical analysis of Embodied Carbon Intensity (ECI) can both refine rating benchmarks and predict EC during the planning stage, enabling early mitigation strategies. This study establishes a systematic baseline analysis workflow for ECI and develops a machine learning (ML) tool for early-stage building design, using Taiwanese residential buildings as a case study. To overcome limited empirical data, Latin Hypercube Sampling (LHS) was applied for data augmentation.
Results show a simulated ECI median of 458.42 kgCO2e/m2, with a mere 2.64% variance relative to the original cases, confirming high statistical representativeness via K-S testing. The ECI distribution of residential buildings in Taiwan exhibits a right-skewed pattern, with the simulated baseline median serving as a crucial reference for industry assessment. Furthermore, carbon-reduction scenario simulations indicate that increasing low-carbon concrete penetration to 80% could reduce the baseline by 3.38% to 442.92 kgCO2e/m2. For policy implementation, this study proposes a "cumulative percentage method" for incentive policies to identify the top 20% low-carbon buildings, and a "confidence interval method" based on industry averages for mandatory regulations. Regarding prediction tools, a Support Vector Regression (SVR) model demonstrated the most stable performance among various ML algorithms, achieving an R2 of 0.760 and keeping the MAPE within 3.5%. This confirms significant correlations between key architectural features and ECI, providing designers with a scientific decision-support tool to estimate ECI ranges during early planning phases accurately.

1.杜怡萱、李雅琪 黃誠中（2016）。RC建築結構體建材用量與碳排放量之影響因子研究。《建築學報》，第95期，59-73頁。
2.黃誠中（2016）。RC中層集合住宅結構體建材用量與碳排放量影響因子研究。碩士論文，國立成功大學。
3.楊忠翰（2022）。建築設計變因對中高層RC建築結構體建材用量與碳排放量之影響。碩士論文，國立成功大學。
4.內政部建築研究所（2023）。綠建築評估手冊-基本型（2023年版）。內政部建築研究所。
5.內政部（2023）。建築物混凝土結構設計規範。內政部（2023）。
6.內政部建築研究所（2024）。建築能效評估手冊（2024年版）。內政部建築研究所。
7.內政部建築研究所（2025）。低碳（低蘊含碳）建築評估手冊（2025年版）。內政部建築研究所。
8.林淑樺（2025）。集合住宅在設計初期之多目的最佳化：生命週期碳排、成本與舒適度。碩士論文，國立成功大學。

1.Allison, P. D. (2009). Missing data. The SAGE handbook of quantitative methods in psychology, 23, 72-89.
2.Bellman, R. E. (2015). Adaptive control processes: a guided tour. Princeton university press.
3.Belsley, D. A., Kuh, E., & Welsch, R. E. (1980). Regression diagnostics: Identifying influential data and sources of collinearity. John Wiley & Sons.
4.Benke, B., Jensen, A., Chafart, M., Simonen, K., & Lewis, M. (2025). Embodied Carbon Benchmark Report: Embodied Carbon Budgets and Analysis of 292 Buildings in the US and Canada. In Carbon Leadership Forum, Apr.
5.Benoit, K. (2011). Linear regression models with logarithmic transformations. London School of Economics, London, 22(1), 23-36.
6.Bessel, F. W. (1813). Untersuchungen über die Bahn des Olberschen Kometen. na.
7.Bouckaert, S., Pales, A. F., McGlade, C., Remme, U., Wanner, B., Varro, L., D'Ambrosio, D., & Spencer, T. (2021). Net zero by 2050: A roadmap for the global energy sector.
8.BRE Global. (2024). BREEAM International New Construction 2024 Technical Manual (SD233). BRE. https://www.breeam.com/
9.Breiman, L. (2001). Random forests. Machine learning, 45(1), 5-32.
10.Cohen, J. (2013). Statistical power analysis for the behavioral sciences. routledge.
11.Collings, D. (2020). Carbon footprint benchmarking data for buildings. The Structural Engineer, 98(11), 10-13. https://doi.org/https://doi.org/10.56330/TFOG9016
12.Cook, R. D. (1977). Detection of Influential Observation in Linear Regression. Technometrics,19(1), 15–18.
13.Cook, R. D., & Weisberg, S. (1982). Residuals and Influence in Regression.
14.Davis, R. A., Lii, K. S., & Politis, D. N. (2011). Remarks on some nonparametric estimates of a density function. In Selected Works of Murray Rosenblatt (pp. 95-100). New York, NY: Springer New York.
15.Draper, N. R., & Smith, H. (1998). Applied regression analysis (Vol. 326). John Wiley & Sons.
16.Drucker, H., Burges, C. J., Kaufman, L., Smola, A., & Vapnik, V. (1996). Support vector regression machines. Advances in neural information processing systems, 9.
17.European Parliament & Council of the European Union. (2024). Directive (EU) 2024/1275 on the energy performance of buildings (recast). EUR-Lex
18.Fisher, R. A. (1970). Statistical methods for research workers. In Breakthroughs in statistics: Methodology and distribution (pp. 66-70). New York, NY: Springer New York.
19.Freund, R. J., & Wilson, W. J. (2003). Statistical methods. Elsevier.
20.Galton, F. (1875). IV. Statistics by intercomparison, with remarks on the law of frequency of error. The London, Edinburgh, and Dublin Philosophical Magazine and Journal of Science, 49(322), 33-46.
21.Galton, F. (1889). Natural Inheritance. London: Macmillan.
22.German Sustainable Building Council (DGNB). (2023). DGNB system for new construction of buildings (Version 2023). https://www.dgnb-system.de/en/
23.Gholamy, A., Kreinovich, V., & Kosheleva, O. (2018). Why 70/30 or 80/20 relation between training and testing sets: A pedagogical explanation. International Journal of Intelligent Technologies and Applied Statistics, 11(2), 105-111.
24.Hair, J. F., Black, W. C., Babin, B. J., & Anderson, R. E. (2019). Multivariate data analysis.
25.Hair, J. F., Hult, G. T. M., Ringle, C. M., Sarstedt, M., Danks, N. P., & Ray, S. (2021). Partial least squares structural equation modeling (PLS-SEM) using R: A workbook. Springer international publishing.
26.Hamilton, I., Kennard, H., Rapf, O., Amorocho, J., Steuwer, S., Kockat, J., & Toth, Z. (2024). 2023 Global Status Report for Buildings and Construction-Beyond foundations: Mainstreaming sustainable solutions to cut emissions from the buildings sector.
27.Han, J., Pei, J., & Tong, H. (2022). Data mining: concepts and techniques(Third Edition). Morgan kaufmann.83-124
28.Hastie, T., Tibshirani, R., & Friedman, J. (2009). The elements of statistical learning: data mining, inference, and prediction.
29.Hoerl, A. E., & Kennard, R. W. (1970). Ridge regression: Biased estimation for nonorthogonal problems. Technometrics, 12(1), 55-67.
30.IEA, P. (2023). Tracking clean energy progress 2023. IEA Paris, France.
31.Jungclaus, M. A., Grant, N., Torres, M. I., Arehart, J. H., & Srubar III, W. V. (2024). Embodied carbon benchmarks of single-family residential buildings in the United States. Sustainable Cities and Society, 117, 105975.
32.Juran, J. M., Gryna, F. M., & Bingham, R. S. (1951). Quality control handbook (Vol. 3, p. 2). New York: McGraw-hill.
33.Kaufman, S., Rosset, S., Perlich, C., & Stitelman, O. (2012). Leakage in data mining: Formulation, detection, and avoidance. ACM Transactions on Knowledge Discovery from Data (TKDD), 6(4), 1-21.
34.Kim, J. H. (2009). Estimating classification error rate: Repeated cross-validation, repeated hold-out and bootstrap. Computational statistics & data analysis, 53(11), 3735-3745.
35.Kohavi, R. (1995, August). A study of cross-validation and bootstrap for accuracy estimation and model selection. In Ijcai (Vol. 14, No. 2, pp. 1137-1145).
36.Kuhn, M., & Johnson, K. (2013). Applied predictive modeling (Vol. 26, p. 13). New York: Springer.
37.Lewis, C. D. (1982). Industrial and business forecasting methods: A practical guide to exponential smoothing and curve fitting. (No Title).
38.Little, R. J., & Rubin, D. B. (2019). Statistical analysis with missing data. John Wiley & Sons.
39.Luo, L., & Cai, W. (2025). An innovative method for rapid estimation and optimization of building embodied carbon emissions. Sustainable Cities and Society, 130, 106556.
40.Luo, L., Chen, Y., Feng, W., Wu, J., & Cai, W. (2025). Progressive prediction of embodied carbon emissions across stages of schematic design with machine learning. Journal of Cleaner Production, 517, 145799.
41.MacLeamy, P. (2004). Collaboration, integrated information and the project lifecycle in building design, construction and operation. WP-1202, The construction users roundtable.
42.Marquardt, D. W. (1970). Generalized inverses, ridge regression, biased linear estimation, and nonlinear estimation. Technometrics, 12(3), 591-612.
43.McKay, M. D., Beckman, R. J., & Conover, W. J. (2000). A comparison of three methods for selecting values of input variables in the analysis of output from a computer code. Technometrics, 42(1), 55-61.
44.Moore, D. S., Notz, W., & Fligner, M. A. (2013). The Basic Practice of Statistics. W.H. Freeman and Company.
45.Ng, W. L., Azmi, A. M., Dahlan, N. Y., & Woon, K. S. (2024). Predicting life cycle carbon emission of green office buildings via an integrated LCA-MLR framework. Energy and Buildings, 316, 114345.
46.One Click, L. C. A. (2021). Embodied carbon benchmarks for European buildings. In: One Click LCA.
47.Pareto, V. (1964). Cours d'économie politique (Vol. 1). Librairie Droz.
48.Paris Agreement (2015). In United Nations Framework Convention on Climate Change (UNFCCC). Retrieved from UNFCCC website.
49.Parzen, E. (1962). On estimation of a probability density function and mode. The annals of mathematical statistics,33(3), 1065-1076.
50.Pearson, K. (1894). III. Contributions to the mathematical theory of evolution. Proceedings of the Royal Society of London, 54(326-330), 329-333.
51.Pearson, K. (1895). VII. Note on regression and inheritance in the case of two parents. Proceedings of the Royal Society of London, 58(347-352), 240-242.
52.Protocol, K. (1997). Kyoto protocol. UNFCCC Website. Available online: http://unfccc. int/kyoto_protocol/items/2830. php (accessed on 1 January 2011), 230-240.
53.Röck, M., Saade, M. R. M., Balouktsi, M., Rasmussen, F. N., Birgisdottir, H., Frischknecht, R., Habert, G., Lützkendorf, T., & Passer, A. (2020). Embodied GHG emissions of buildings–The hidden challenge for effective climate change mitigation. Applied energy, 258, 114107.
54.Saltelli, A., Ratto, M., Andres, T., Campolongo, F., Cariboni, J., Gatelli, D., ... & Tarantola, S. (2008).Global sensitivity analysis: the primer. John Wiley & Sons.
55.Sarker, I. H. (2021). Machine learning: Algorithms, real-world applications and research directions. SN computer science, 2(3), 1-21.
56.Silverman, B. W. (1986). Density estimation for statistics and data analysis. Routledge.
57.Smola, A. J., & Schölkopf, B. (2004). A tutorial on support vector regression. Statistics and computing, 14(3), 199-222.
58.Snedecor, G. W., & Cochran, W. G. (1989). Statistical methods, 8thEdn. Ames: Iowa State Univ. Press Iowa, 54, 71-82.
59.Standardization), C. E. C. f. (2011). Sustainability of construction works—Assessment of environmental performance of buildings—Calculation method. In (Vol. EN 15978:2011).
60.Stone, M. (1974). Cross‐validatory choice and assessment of statistical predictions. Journal of the royal statistical society: Series B (Methodological), 36(2), 111-133.
61.Student. (1908). The probable error of a mean. Biometrika, 1-25.
62.Tibshirani, R. (1996). Regression shrinkage and selection via the lasso. Journal of the Royal Statistical Society Series B: Statistical Methodology, 58(1), 267-288.
63.Torres, M., & Srubar III, W. V. (2025). Characterizing statistical uncertainty and variability of building material emissions in probabilistic whole-building life cycle assessment using kernel density estimation. Building and Environment, 113442.
64.Toth, Z., Broer, R., Graaf, L., & Röck, M. (2024). How to establish Whole Life Carbon benchmarks: Insights and lessons learned from emerging approaches in Ireland. Czechia and Spain.
65.Tukey, J. W. (1977). Exploratory data analysis (Vol. 2, pp. 131-160). Reading, MA: Addison-wesley.
66.United Nations (1992). United Nations Framework Convention on Climate Change. New York: UN.
67.U.S. Green Building Council (2019). LEED v4.1 Building Design and Construction.
68.Vapnik, V. N. (1999). An overview of statistical learning theory. IEEE transactions on neural networks, 10(5), 988-999.
69.Willmott, C. J., & Matsuura, K. (2005). Advantages of the mean absolute error (MAE) over the root mean square error (RMSE) in assessing average model performance. Climate research, 30(1), 79-82.
70.Wiik, M. K. (2025). Developing whole-life carbon benchmark values for Norwegian buildings. Building Research & Information, 53(3), 345–358.
71.Wright, S. (1921). Correlation and causation. Journal of agricultural research, 20(7), 557.
72.Zhang, X., Li, Y., Chen, H., Yan, X., & Liu, K. (2023). Characteristics of embodied carbon emissions for high-rise building construction: A statistical study on 403 residential buildings in China. Resources, Conservation and Recycling, 198, 107200. https://doi.org/https://doi.org/10.1016/j.resconrec.2023.107200
73.Zhao, S., Deng, H., Cao, J., & Gustaf, M. (2025). Digital transformation of construction enterprises and carbon emission reduction: evidence from listed companies. Frontiers in Environmental Science, 13. https://doi.org/10.3389/fenvs.2025.1570182