The synergetic reduction of CO₂ and PM_2.5 emissions has received much attention in China in recent years. A comprehensive evaluation of the synergy between CO₂ emission reduction (CER) and PM_2.5 emission reduction (PER) would provide valuable information for developing synergetic control policies. Thus, we constructed a comprehensive CO₂-PM_2.5-emission-reduction index system and evaluated the synergy between CER and PER, using the coupling coordination degree (CCD) and relative development degree (RDD) model in China’s 329 cities from 2003 to 2017. The spatiotemporal characteristics of the CCD were analyzed on the national, regional, and urban scales. Furthermore, we used the spatial autocorrelation analysis, kernel density estimation, and Dagum Gini coefficient to investigate the spatial autocorrelation, evolutionary characteristics, and regional differences of the CCD. The results indicate that (1) the synergy between CO₂ and PM_2.5 emissions’ reductions showed an upward trend, and the lowest CCD values occurred in NW and Shanghai on the regional and urban scales, respectively; (2) the CCD showed obvious spatial clustering characteristics, with 75% of the cities located in the “High–High” or “Low–Low” clustering zones in the Moran scatter plots in 2017; (3) the polarization of CCD in SC, MYR, and SW showed intensified trends; (4) and the hypervariable density was the largest contributor to the overall difference in the CCD. Our findings suggest that more attention should be paid to the top-level design of the policies, technological innovation, and cross-regional or intercity cooperation.

This study has made significant findings in A-share market data processing and portfolio management. Firstly, by adopting the Lasso method and CPCA framework, we effectively addressed the problem of multicollinearity among feature indicators, with the Lasso method demonstrating superior performance in handling this issue, thus providing a new method for financial data processing. Secondly, Deep Feedforward Neural Networks (DFN) exhibited exceptional performance in portfolio management, significantly outperforming other evaluated machine learning methods, and achieving high levels of out-of-sample performance and Sharpe ratios. Additionally, we consistently identified price changes, earnings per share, net assets per share, and excess returns as key factors influencing predictive signals. Finally, this study combined the Lasso method with DFN, providing a new perspective and methodological support for asset pricing measurement in the financial field.