Carbon Price Forecasting Based on a Hybrid Model of Fractal Interpolation and Machine Learning
The carbon emission trading market is one of the core tools for achieving the goals of carbon peak and carbon neutrality.Predicting its market fluctuations is of great significance for the pro-duction entities to have stable expectations and promote carbon peaking.Machine learning algorithms are used to predict interpolation points,and then the original iterative function system is extended for fractal interpolation for hybrid forecasting.SVM,RF,and LSTM are respectively hybridized and em-pirically analyzed based on the data from the Guangzhou Carbon Exchange.The results show that the hybrid model is superior to traditional interpolation algorithms and is more suitable for non-stationary financial time series such as carbon prices.
carbon trading marketfractal interpolationmachine learningshort-term time se-ries forecasting
万方数据
维普
