The impact mechanism and prediction of digital economy on carbon emissions——Construction of LSTM model based on double input features

Abstract

This paper constructs a digital economy index measurement system including four dimensions : digital infrastructure, digital innovation environment, digital industry scale and digital integration degree. Combined with the dual-input feature LSTM deep learning model, this paper systematically studies the development level of digital economy and its impact on carbon emissions in various provinces of China. The study finds that the development of China 's digital economy shows significant regional heterogeneity. The eastern region has a high index and outstanding technological innovation advantages, the central region is relatively balanced and the average annual growth rate is stable, and the western region has significant growth potential. Through Pearson correlation coefficient verification, the digital economy index is significantly negatively correlated with carbon emission intensity. By verifying the substantial impact of the digital economy on carbon emission projections, it provides a scientific basis for differentiated carbon emission reduction policies.