Based on the theory of cloud control systems, an intelligent transportation cyber-physical cloud control system is designed due to the problems of complex objects, big data, high demand for transmission and calculation and poor real-time control ability in the modern intelligent transportation cyber-physical network. It includes intelligent transportation edge control technology and intelligent transportation network virtualization technology. Based on the big data of intelligent traffic flow, two intelligent learning methods, deep learning and extreme learning machine, are used to train and predict the traffic flow data on the servers of the cloud control management center. The short time traffic flow and the congestion of roads are predicted accurately. Then the real-time traffic flow control strategy is obtained by intelligent optimization scheduling algorithm in the cloud. The problem of traffic flow distribution in congested roads is solved and the dynamic performance of intelligent transportation control systems can be improved.

After the 2008 financial crisis, companies in China begun paying more attention to sustainable development. This article attempts to examine whether and how digital transformation affects total factor productivity (TFP) at the firm-level. Using 2913 listed companies in China from 2012 to 2018, this study finds that digital transformation is positively associated with corporate TFP in China. Our explanatory variable of firm-level digitalization index is constructed via text analysis methods. After a series of robustness checks and different attempts that mitigate endogeneity concerns, our findings remain valid. However, traditional information and communication technologies (ICT) cannot improve corporate TFP. Further analyses of three plausible channels indicate that digital technologies improve TFP primarily through cost reduction and human substitution rather than supervision advantage. The results indicate that firms achieve actual benefits from the digital transformation, and how digital transformation improve the sustainable development. This study could serve as a policy inspiration for other developing countries.

As global climate change intensifies and carbon emission policies become increasingly stringent, carbon risk has emerged as a crucial factor influencing corporate operations and financial markets. Based on data from A-share listed companies in China from 2009 to 2022, this paper empirically examines the pricing mechanism of carbon risk in the Chinese capital market and explores how different corporate signaling behaviors affect the carbon risk premium. The findings reveal the following: (1) Carbon risk exhibits a significant positive premium (annualized at about 1.33% per standard deviation), which remains robust over longer time windows and after replacing the measurement variables. (2) Heterogeneity analysis shows that the carbon risk premium is not significant in high-energy-consuming industries or before the signing of the Paris Agreement, possibly due to changes in investor expectations and increased green awareness. Additionally, a significant difference in the carbon risk premium exists between brown and green stocks, reflecting a “labeling effect” of green attributes. (3) Issuing green bonds, as an active corporate signaling behavior, effectively mitigates the carbon risk premium, indicating that market investors highly recognize and favor firms that actively convey green signals. (4) A “greenwashing” indicator constructed from textual analysis of environmental information disclosure suggests that greenwashing leads to a mispricing of the carbon risk premium. Companies that issue false green signals—publicly committing to environmental protection but failing to implement corresponding emission reduction measures—may mislead investors and create adverse selection problems. Finally, this paper provides recommendations for corporate carbon risk management and policy formulation, offering insights for both research and practice in the field.

Interpretability analysis methods, such as LIME and SHAP, are widely employed to explain the predictions of artificial intelligence models; however, they primarily function as post hoc tools and do not directly quantify the intrinsic interpretability of the models. Although it is commonly assumed that model transparency decreases with increasing complexity, there is currently no standardized framework for evaluating interpretability as an inherent property of AI models. In this study, we examine the prediction of bond defaults using several widely used machine learning algorithms. The classification performance of each algorithm is first evaluated, followed by the application of LIME and SHAP to assess the influence of input features on model outputs. Based on these analyses, we propose a novel approach for quantifying intrinsic model interpretability. The results align with theoretical expectations and provide insights into the trade-off between model complexity and interpretability.

To establish a micro foundation to understand the impacts of greenhouse gas (GHG) emission regulations and financial development levels on firms’ GHG emissions, we build a two-stage dynamic game model to incorporate GHG emission regulations (in terms of an emission tax) and financial development (represented by the corresponding financing cost) into a two-echelon supply chain. With the subgame perfect equilibrium, we identify the conditions to determine whether an emission regulatory policy and/or financial development can affect GHG emissions in the supply chain. We also reveal the impacts of the strictness of GHG emission regulation, the financial development level, and the unit GHG emission rate on the operations of the supply chain and the corresponding profitability implications. Managerial insights are also discussed.

Since the outbreak of credit risk, researching on corporate credit rating has been brought into investors, the government and scholars focus. This paper constructs an optimal K-means clustering Group Method of Data Handling model can effectively improve the accuracy of rating results, reduce the computational complexity, and this paper proves the model under the least squares estimation can get the optimal results. This article uses Chinese corporate credit rating and financial indexes to study, comparing its results with the concequences of Hidden Markov GMDH model and other traditional neural network models. The empirical outcomes show that the K-means clustering GMDH model is better than Hidden Markov GMDH model and the remaining four neural network models, indicating that the method can effectively improve the accuracy of corporate credit rating assessment and reduce the cost of rating.