A combined forecasting framework including point prediction and interval prediction for carbon emission trading prices

With the severe situation of climate and environmental issues, carbon emissions have aroused great attention from the academic community and industry. Building and improving the carbon trading market has become the focus, among which the prediction of carbon emissions trading prices is crucial. However, randomness and instability make it a challenging task to predict price series accurately. To obtain accurate point prediction results, a combined prediction idea is constructed in the study. In addition, considering that the point prediction framework contains less data information, in order to bridge this gap, an interval prediction frame is developed in this research. The optimal distribution of data sequence is obtained by using distribution function and optimization techniques, and successfully achieve different levels of uncertainty prediction according to the point prediction results. Several comparison experiments were carried out using the daily price of carbon emission futures of the European Union Emissions Trading System and the forecasting performance of the proposed forecasting framework was verified through experimental analysis. Experiments and discussion demonstrate the superior performance of the proposed prediction scheme and its excellent forecasting ability for carbon emissions trading price prediction.

Automated summary

With the severe situation of climate and environmental issues, predicting carbon emissions trading prices has become crucial. This study proposes a combined prediction idea and develops an interval prediction framework to achieve different levels of uncertainty prediction. The experimental analysis verifies the superior performance and excellent forecasting ability of the proposed prediction scheme for carbon emissions trading price prediction.