Application of Digital Economy Machine Learning Algorithm for Predicting Carbon Trading Prices Under Carbon Reduction Trends

Abstract

Due to the increasing demand for fossil fuels, excessive emissions of greenhouse gases such as CO2 have been caused. With the intensification of global climate anomalies and warming, how to reduce greenhouse gas emissions is an important issue currently facing the international community. The influencing factors of carbon price are complex, and accurate prediction of carbon price is a difficult problem. There are still some problems in the existing carbon trading price prediction models, such as insufficient understanding of the enormous potential of machine learning models to ilift the performance. The study will use two machine learning models that can address the shortcomings of traditional artificial intelligence models as the basic prediction models. The specific content includes machine learning prediction models that extend to extreme learning machine theory and fuzzy inference system theory. By integrating data preprocessing algorithms, artificial intelligence optimization algorithms, feature selection algorithms, etc., this study constructs and applies a carbon trading price prediction model from multiple perspectives to compensate for the shortcomings in current research. The corresponding values for each indicator in the algorithm are 5.6214E-12 (maximum), 2.8546E-12 (minimum), 4.0239E-12 (mean), and 5.4402E-13 (variance). Compared with other comparative optimization algorithms, this indicates that the hybrid optimization algorithm is an efficient optimization method for the model, which can effectively optimize different problems. In theory, the proposed multiple improved carbon trading price prediction models can theoretically compensate for the shortcomings in existing carbon trading price predictions.