Credit card fraud detection is a critical challenge in the financial sector due to the rapidly evolving tactics of fraudsters and the significant class imbalance betweenegitimate and fraudulent transactions. Traditional models, while effective to some extent, often suffer from high false positive rates and fail to generalize well to emerging fraud patterns. In this paper, we propose a novel approach that integrates a Mixture of Experts (MoE) model with a Deep Neural Network-based Synthetic Minority Over-sampling Technique (DNN-SMOTE) to enhance fraud detection performance. The MoE modeleverages multiple specialized expert networks, each trained to detect specific types of fraud, while the DNN-SMOTE generates high-quality synthetic samples to address the class imbalance. Our experimental results on a publicly available dataset demonstrate that the proposed method achieves a classification accuracy of , a true positive rate of , and a true negative rate of . The Matthews Correlation Coefficient (MCC) of 0.7883 further highlights the model’s balanced performance in detecting fraudulent transactions. These results underscore the effectiveness of combining MoE with DNN-SMOTE, offering a robust solution for real-world credit card fraud detection scenarios.
Drawing on a comprehensive sample of China’s A-share listed companies spanning 2007–2022, this paper provides empirical evidence on how institutional investment horizons shape corporate green innovation, with special attention to the operative mechanisms. After implementing standard data screening protocols, the resulting dataset comprising 24,362 firm-year observations is examined through panel regression techniques. The results demonstrate that longer investment horizons of institutional investors significantly promote corporate green innovation. Our mechanism analysis indicates that environmental regulations and subsidy policies significantly moderate the relationship between institutional investors’ investment horizons and green innovation. These findings provide both a novel theoretical perspective and robust empirical evidence on the determinants of corporate green innovation, and yield valuable policy insights for leveraging capital markets to facilitate the corporate transition to sustainability. It also provides specific references for policymakers to facilitate corporate green transformation and support the dual-carbon goals by guiding and incentivizing greater influence exerted by long-term institutional investors through corporate governance mechanisms.
Compound heat–drought (CHD) exposure poses a growing risk to grain production safety under climate change. Using a panel of Chinese cities from 2010 to 2024, we estimate its effects on grain yield anomalies. We also test whether the compound effect exceeds the sum of heat-only and drought-only exposure. Greater CHD exposure shifts grain yield anomalies toward more negative values, indicating larger losses relative to the long-run trend. This relationship remains robust across alternative CHD definitions, alternative measures of production anomalies, and adjusted samples. The adverse effects are stronger in major grain-producing regions and areas with low topographic relief. Mediation analysis indicates that CHD exposure worsens grain yield anomalies partly by intensifying soil moisture deficits. Irrigation capacity, agricultural technological innovation, agricultural insurance coverage, and agricultural organizational capacity mitigate these effects. They operate through resource provision, technological adaptation, risk sharing, and organizational coordination. The adverse effect of CHD exposure also exceeds the additive effects of heat-only and drought-only exposure, indicating synergistic damage amplification. These findings show how compound climate extremes generate grain yield anomalies. They also highlight the need to strengthen agricultural adaptive capacity and tailor climate risk management policies to safeguard food security.