Scholarlink

The carbon emissions trading system plays a crucial role in regulating greenhouse gas emissions and promoting sustainable practices. Accurate forecasting of carbon trading prices is essential for optimizing market strategies and supporting climate goals. This study introduces a novel hybrid approach that combined complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) and extreme gradient boosting (XGBoost) for the prediction of carbon prices. The CEEMDAN method decomposes time-series data into multiple intrinsic mode functions (IMFs), which are then input into the XGBoost model for forecasting. The performance of the CEEMDAN-XGBoost model is compared with traditional models such as ARIMA and ARIMA-LSTM. The results demonstrate a significant improvement in forecast accuracy, achieving an improvement in 48% performance as measured by RMSE, MAE, and MAPE.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Forecasting Carbon Emissions Trading Prices in Asia Using a CEEMDAN-XGBoost Hybrid Model

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Ouyang Mutian,
  • Aung Pyae,
  • Yang Hanlin,
  • J. Joshua Thomas

摘要

The carbon emissions trading system plays a crucial role in regulating greenhouse gas emissions and promoting sustainable practices. Accurate forecasting of carbon trading prices is essential for optimizing market strategies and supporting climate goals. This study introduces a novel hybrid approach that combined complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) and extreme gradient boosting (XGBoost) for the prediction of carbon prices. The CEEMDAN method decomposes time-series data into multiple intrinsic mode functions (IMFs), which are then input into the XGBoost model for forecasting. The performance of the CEEMDAN-XGBoost model is compared with traditional models such as ARIMA and ARIMA-LSTM. The results demonstrate a significant improvement in forecast accuracy, achieving an improvement in 48% performance as measured by RMSE, MAE, and MAPE.