隨著全球氣候變化議題受到廣泛關注，尋找有效的減碳解決方案成為國際社會的共識，碳排放交易成為有效調控溫室氣體排放的市場機制之一。碳權金融商品作為一種創新的市場機制，通過為碳排放定價，激勵企業和個人減少溫室氣體排放，從而成為應對氣候變化的重要工具。本研究採用並評估了ARIMA統計模型以及四種機器學習技術，包括分類與迴歸樹CART、輕量級梯度提升機LightGBM、隨機森林RF、極限梯度提升機XGBoost，旨在對碳權金融商品的時間序列數據進行精確預測。通過深入分析，本研究概述了各種預測方法的效能，目的在於識別哪一種算法能夠提供最為準確的預測結果。以碳排期貨價格預測為例，並以烏俄戰爭作為大事件，查看戰爭前後預測準確度有沒有明顯改變。
    實證結果顯示，ARIMA統計模型預測效果比四種機器學習效果來得好，並且可以對碳排期貨金融建立一個有效的工具。

    As global climate change issues gain widespread attention, the search for effective carbon reduction solutions has become a consensus in the international community. Carbon emission trading has emerged as one of the effective market mechanisms for regulating greenhouse gas emissions. As an innovative market mechanism, carbon credit financial products incentivize enterprises and individuals to reduce greenhouse gas emissions by setting a price on carbon, thereby becoming a crucial tool in addressing climate change. This study employs and evaluates the ARIMA statistical model along with four machine learning techniques: Classification and Regression Trees (CART), Light Gradient Boosting Machine (LightGBM), Random Forest (RF), and Extreme Gradient Boosting (XGBoost), aiming to accurately predict the time series data of carbon credit financial products. Through thorough analysis, this research outlines the performance of various predictive methods, with the objective of identifying which algorithm can deliver the most accurate results. Taking the prediction of carbon futures prices as a case study, and considering the Russia-Ukraine war as a significant event, the study examines whether there is a noticeable change in prediction accuracy before and after the war.
    The empirical results show that the ARIMA statistical model performs better than the four machine learning techniques and can serve as an effective tool for managing carbon futures finances.

目錄
中文摘要	I
英文摘要	III
目錄	V
圖目錄	VII
表目錄	IX
第一章  緒論	1
第一節 研究背景與動機	1
第二節 研究目的	5
第三節 研究流程	6
第二章  文獻探討	7
第一節 碳權金融商品	7
第二節 機器學習之應用	10
第三節 時間序列之應用	12
第三章 研究方法	16
第一節 研究架構	16
第二節 資料來源	17
第三節 研究模型	18
3.3.1 ARIMA	19
3.3.2 CART	20
3.3.3 LightGBM	22
3.3.4 RF	25
3.3.5 XGBoost	27
第四節 模型評估指標	29
第四章  實證結果	32
第一節 模式建構	32
第二節 最適滯後期數	34
第三節 模型預測結果	37
4.3.1碳排期貨預測結果	37
4.3.2烏俄戰爭前後碳排期貨預測結果	40
4.3.3穩健性評估	45
4.3.4綜合比較	52
第五章  結論與建議	53
第一節 研究結論	53
第二節 未來研究方向與建議	54
參考文獻	56



圖目錄
圖1-1 研究流程圖	6
圖3-1 研究架構圖	17
圖3-2 碳排放期貨價格歷史走勢圖	18
圖 3-3 ARIMA虛擬碼	20
圖3-4 CART虛擬碼	22
圖3-5 LIGHTGBM虛擬碼	25
圖3-6 RF虛擬碼	27
圖3-7 XGBOOST虛擬碼	28
圖4-1 訓練集與測試集比例圖(80/20)	33
圖4-2 ARIMA(80/20)測試集預測	39
圖4-3 CART(80/20)測試集預測	39
圖4-4 LIGHTGBM(80/20)測試集預測	39
圖4-5 RF(80/20)測試集預測	40
圖4-6 XGBOOST(80/20)測試集預測	40
圖4-7 烏俄戰爭前期貨價格趨勢	41
圖4-8 烏俄戰爭後期貨價格趨勢	41
圖4-9 本研究碳排放期貨資料及事件期間時間軸	42
圖4-10 各方法於(90/10)測試集預測圖	47
圖4-11 各方法於(70/30)測試集預測圖	48
圖4-12 各方法於(60/40)測試集預測圖	49



表目錄
表4-1 ARIMA滯後期數	35
表4-2 CART滯後期數	36
表4-3 LIGHTGBM滯後期數	36
表4-4 RF滯後期數	36
表4-5 XGBOOST滯後期數	36
表4-6 五個方法最適滯後期數之綜合比較	37
表4-7 碳排期貨預測結果於80/20資料切分比例比較	38
表4-8 80/20烏俄戰爭前後模型評估指標	45
表4-9 碳排期貨預測結果於90/10資料切分比例比較	46
表4-10 碳排期貨預測結果於70/30資料切分比例比較	47
表4-11 碳排期貨預測結果於60/40資料切分比例比較	48
表4-12 90/10烏俄戰爭前後模型評估指標	50
表4-13 70/30烏俄戰爭前後模型評估指標	51
表4-14 60/40烏俄戰爭前後模型評估指標	51

參考文獻
英文文獻
Alberola, E., Chevallier, J., & Chèze, B. (2008). The EU Emissions Trading Scheme: Disentangling the effects of industrial production and CO2 emissions on carbon prices. EconomiX Working Papers, 2008-12. University of Paris Nanterre, EconomiX.
Ahmed, N. K., Atiya, A. F., Gayar, N. E., & El-Shishiny, H. (2010). An empirical comparison of machine learning models for time series forecasting. Econometric reviews, 29(5-6), 594-621.
Anderson, G., & Pfahringer, B. (2009). Random relational rules (Doctoral dissertation). University of Waikato, Hamilton, New Zealand.
Buitinck, L., Louppe, G., Blondel, M., Pedregosa, F., Mueller, A., Grisel, O., Niculae, V., Prettenhofer, P., Gramfort, A., Grobler, J., Layton, R., VanderPlas, J., Joly, A., Holt, B. & Varoquaux, G. API design for machine learning software: experiences from the scikit-learn project. ECML PKDD Workshop: Languages for Data Mining and Machine Learning. 2013, 108–122.
Breiman, L. (2001). Random forests. Machine learning, 45, 5-32.
Boonamnuay, S., Kerdprasop, N., & Kerdprasop, K. (2018). Classification and regression tree with resampling for classifying imbalanced data. International Journal of Machine Learning and Computing, 8(4), 336-340.
Chen, T., & Guestrin, C. (2016, August). Xgboost: A scalable tree boosting system. In Proceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining (pp. 785-794).
Coase, R. H. (1960). The problem of social cost. Journal of Law and Economics, 3, 1-44.
Cordesman, A. H. (2023). Lasting Strategic Impact of the War in Ukraine. Center for Strategic and International Studies (CSIS).
Congressional Research Service. (2022). Russia’s invasion of Ukraine: Overview and issues for Congress. Retrieved from https://crsreports.congress.gov/product/pdf/IF/IF12040
Durbin, J., & Koopman, S. J. (2012). Time series analysis by state space methods (Vol. 38). OUP Oxford.
Dunbray, N., Rane, R., Nimje, S., Katade, J., & Mavale, S. (2021, October). A Novel Prediction Model for Diabetes Detection Using Gridsearch and A Voting Classifier between Lightgbm and KNN. In 2021 2nd Global Conference for Advancement in Technology (GCAT) (pp. 1-7). IEEE.
Friedman, J. H. (2001). Greedy function approximation: a gradient boosting machine. Annals of statistics, 1189-1232.
Gao, Y., Mosalam, K. M., Chen, Y., Wang, W., & Chen, Y. (2021). Auto-regressive integrated moving-average machine learning for damage identification of steel frames. Applied Sciences, 11(13), 6084. 
International Carbon Action Partnership. (2021). Emissions trading worldwide: Status report 2021. Retrieved March 26, 2021, from https://icapcarbonaction.com/zh/?option=com_attach&task=download&id=683
Liu, Z., Deng, Z., Davis, S. J., Giron, C., & Ciais, P. (2022). Monitoring global carbon emissions in 2021. Nature Reviews Earth & Environment, 3(4), 217-219.
Mansanet-Bataller, M., Pardo, A., & Valor, E. (2007). CO2 prices, energy and weather. The energy journal, 28(3), 73-92.
Keppler, J. H., & Mansanet-Bataller, M. (2010). Causalities between CO2, electricity, and other energy variables during phase I and phase II of the EU ETS. Energy policy, 38(7), 3329-3341. 
Kobiela, D., Krefta, D., Król, W., & Weichbroth, P. (2022). ARIMA vs LSTM on NASDAQ stock exchange data. Procedia Computer Science, 207,3836-3845.

Mansanet-Bataller, M., & Soriano, P. (2009). The role of CO2 prices in the European Union Emissions Trading Scheme: Evidence from Phase I. [Journal Name], [Volume(Issue)], [Page Range].
Makridakis, S., Spiliotis, E., & Assimakopoulos, V. (2018). Statistical and Machine Learning forecasting methods: Concerns and ways forward. PloS one, 13(3), e0194889. 
Makridakis, S., Wheelwright, S. C., & Hyndman, R. J. (2008). Forecasting methods and applications. John wiley & sons.
McKinsey & Company. (2023). Carbon markets and the path to net-zero emissions. Retrieved from https://www.mckinsey.com/business-functions/sustainability/our-insights/putting-carbon-markets-to-work-on-the-path-to-net-zero
Nerlinger, M., & Utz, S. (2022). The impact of the Russia-Ukraine conflict on energy firms: A capital market perspective. Finance Research Letters, 50, 103243. 
Natarajan, Y., Wadhwa, G., Sri Preethaa, K. R., & Paul, A. (2023). Forecasting carbon dioxide emissions of light-duty vehicles with different machine learning algorithms. Electronics, 12(10), 2288. 
Ozili, P. K. (2022). Green finance research around the world: a review of literature. International Journal of Green Economics, 16(1), 56-75. 
Orhan, M. A. (2022). The effects of the Russia-Ukraine war on global trade. Economics Bulletin, 42(2), 1234-1256. 
Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M., Prettenhofer, P., Weiss, R., Dubourg, V., Vanderplas, J., Passos, A., Cournapeau, D., Brucher, M., Perrot, M. & Duchesnay, E. Scikit-learn: Machine Learning in Python. Journal of Machine Learning Research. 2011, 12, 2825–2830.
Prohorovs, A. (2022). Russia’s war in Ukraine: Consequences for European countries’ businesses and economies. Journal of risk and financial management, 15(7), 295. 
Quinlan, J. R. (1986). Induction of decision trees. Machine learning, 1, 81-106.
Qin, C., Zhang, Y., Bao, F., Zhang, C., Liu, P., & Liu, P. (2021). XGBoost optimized by adaptive particle swarm optimization for credit scoring. Mathematical Problems in Engineering, 2021(1), 6655510.
Said, S. E., & Dickey, D. A. (1985). Hypothesis testing in ARIMA (p, 1, q) models. Journal of the American Statistical Association, 80(390), 369-374.
Su, L., Yu, W., & Zhou, Z. (2023). Global trends of carbon finance: A bibliometric analysis. Sustainability, 15(8), 6784.
Sokhanvar, A., & Bouri, E. (2022). Commodity price shocks related to the war in Ukraine and exchange rates of commodity exporters and importers. Borsa Istanbul Review, 23(1), 44-54.
Shahzad, U., Sengupta, T., Rao, A., & Cui, L. (2023). [Title of the Article]. [Journal Name], [Volume(Issue)], [Page Range]. [DOI or URL if available].
Van Rossum, G., & Drake, F. L. Python 3 Reference Manual. Scotts Valley. 2009, CA: CreateSpace.
Wood, D., & Dasgupta, B. (1996). Classifying trend movements in the MSCI USA capital market index—A comparison of regression, arima and neural network methods. Computers & Operations Research, 23(6), 611-622.
Zhang, Y. J., Wang, A. D., & Tan, W. (2015). The impact of China's carbon allowance allocation rules on the product prices and emission reduction behaviors of ETS-covered enterprises. Energy Policy, 86, 176-185.
Zhou, J., Huo, X., Xu, X., & Li, Y. (2019). Forecasting the carbon price using extreme-point symmetric mode decomposition and extreme learning machine optimized by the grey wolf optimizer algorithm. Energies, 12(5), 950.
Zhang, L., Lin, J., Qiu, R., Hu, X., Zhang, H., Chen, Q., ... & Wang, J. (2018). Trend analysis and forecast of PM2. 5 in Fuzhou, China using the ARIMA model. Ecological indicators, 95, 702-710.
網路文獻
Investing.com. Carbon Emissions Futures Historical Data. (2024, March 12). 
    檢索自: https://www.investing.com/commodities/carbon-emissions-historical-data
IPCC. (2023). Climate change 2023: Synthesis report.  
　　檢索自：https://www.ipcc.ch/report/sixth-assessment-report-cycle/
National Resources Defense Council. (2024). What causes climate change? 檢索自：https://www.nrdc.org/stories/what-causes-global-warming　
United Nations Framework Convention on Climate Change. (2024, May              30). In Wikipedia, The Free Encyclopedia.
　　檢索自：https://reurl.cc/mMQL1Y
United Nations Framework Convention on Climate Change. (2024). Kyoto Protocol. 
　　檢索自：https://unfccc.int/kyoto_protocol
World Bank. (2022, August 17). What you need to know about results-based climate finance.
    檢索自：https://reurl.cc/DjpgW5