🤖 AI Summary
Addressing the challenge of identifying industry carbon emission drivers under multifactorial influences and high multicollinearity, this study proposes a novel synergistic modeling framework integrating DBSCAN-based unsupervised feature clustering with Lasso and Ridge penalized regression—enabling, for the first time, objective grouping of collinear variables and quantitative attribution of driving factors. Leveraging multi-source energy consumption data across 46 Chinese industries from 2000 to 2019, the method identifies 16 distinct industrial emission patterns and precisely quantifies both their emission characteristics and the relative strengths of dominant drivers. By decoupling collinearity-induced estimation bias and enhancing structural interpretability, the framework overcomes the explanatory limitations of conventional regression in complex systems. It significantly improves the robustness and transparency of driver identification, thereby providing a data-driven, scientifically rigorous foundation for designing differentiated and precision-oriented carbon mitigation policies.
📝 Abstract
This study presents a general analytical framework using DBSCAN clustering and penalized regression models to address multifactor problems with structural complexity and multicollinearity issues, such as carbon emission issue. The framework leverages DBSCAN for unsupervised learning to objectively cluster features. Meanwhile, penalized regression considers model complexity control and high dimensional feature selection to identify dominant influencing factors. Applying this framework to analyze energy consumption data for 46 industries from 2000 to 2019 identified 16 categories in the sample of China. We quantitatively assessed emission characteristics and drivers for each. The results demonstrate the framework's analytical approach can identify primary emission sources by category, providing quantitative references for decision-making. Overall, this framework can evaluate complex regional issues like carbon emissions to support policymaking. This research preliminarily validated its application value in identifying opportunities for emission reduction worldwide.