The environmental Kuznets curve hypothesis: an ML approach to assessing economic growth and environmental sustainability using artificial neural network
Yunqiu Sun,Zhiyu Sun,Zhiman Jiang
DOI: https://doi.org/10.1007/s00500-024-09633-y
IF: 3.732
2024-01-28
Soft Computing
Abstract:In recent years, environmental concerns have emerged as significant obstacles to economic advancement, leading to a critical examination of the complex relationship between environmental sustainability and economic expansion. With the economy expanding swiftly and environmental hazards intensifying, immediate action must be taken to address these problems. This necessitates innovative policies and practices that foster sustainable economic growth while minimizing environmental impact. To solve these issues, this study proposed an extensive analytical framework employing machine learning techniques, particularly artificial neural networks (ANNs), to investigate the Environmental Kuznets Curve (EKC) hypothesis. According to the EKC hypothesis, environmental degradation worsens before economic growth occurs. This study examines GDP per capita and key environmental indicators to measure economic development, using data from Anhui Province, China, from 2012 to 2019. The analysis provides empirical support for the theory and establishes a modeling framework that seamlessly integrates ML methodologies with EKC analysis, systematically examining and assessing the EKC's reliability for Anhui Province. The intricate interplay between Anhui Province's environment and economy, utilizing industrial waste gas, wastewater, and solid waste as critical indicators, evaluates environmental quality. At the same time, per capita GDP is the economic growth parameter. The findings indicate that the EKC does not adhere to an unchanging law but instead displays variations in distribution characteristics. The experimental results of the proposed ANN model demonstrate an impressive accuracy rate of 87%, outperforming five alternative models, including RF, KNN, DT, and LR. This superior model performance underscores the effectiveness of the ANN-based approach in capturing the intricate dynamics linking economic development and environmental pollution.
computer science, artificial intelligence, interdisciplinary applications