Search Results (63)

Machine learning-based credit risk models are commonly assessed using discrimination metrics alone. Such evaluation, however, does not fully capture economic consequences, temporal deployment conditions, or changes in the underlying risk environment. This study develops a decision-aligned, temporally consistent evaluation framework for real-world deployment. Using loan-level data, model performance is examined under a rolling forward validation scheme. A coverage-based alert policy is implemented to reflect operational resource constraints. Predictive discrimination is measured using PR-AUC, while economic performance is evaluated through a cost-sensitive saving function. The false-negative cost parameter (λ) is varied between 5 and 25 to assess sensitivity. Performance is also compared across high- and low-default regimes, and alternative alert budgets of 5%, 10%, and 20% are considered to examine policy stability. The results indicate no systematic decline in PR-AUC over time. Changes in λ do not alter predictive ranking, although economic returns scale proportionally with the cost parameter. Economic gains are higher in high-default regimes, yet no structural deterioration is observed in low-default periods. Increasing coverage improves recall but reduces economic benefit due to higher false-positive costs. To consolidate these stability dimensions, the Unified Policy Stability Index (UPSI) is proposed as a composite measure integrating predictive variability, economic volatility, and regime differences. The index indicates sustained overall stability during the study period. The findings suggest that credit risk model evaluation should extend beyond accuracy-centered metrics and incorporate decision consistency, temporal robustness, and policy stability within a deployment-oriented framework. Full article

Legal element recognition, which identifies discrete factual elements in Chinese court judgments to support judicial analysis and case retrieval, faces a severe long-tail challenge: head-to-tail label-frequency ratios exceed 100:1, and over 60% of sentences carry no label, starving rare elements of training signal. Static reweighting methods assign fixed weights prior to training and cannot respond to the model’s evolving confidence; sample-level meta-learning couples all co-occurring label gradients to a single scalar, preventing independent tail-label amplification. We propose BML-Trans, a boundary-aware meta-learning framework that addresses both limitations. A label-wise meta-weighting mechanism maintains per-label gradient weights updated via bilevel hypergradient descent, decoupling tail-label amplification from co-occurring head labels. A boundary-aware meta-set concentrates calibration signal on high-uncertainty, tail-triggering sentences rather than on easy negatives, and a lightweight Multi-Scale Adapter sharpens the warm-up probability estimates on which boundary selection depends. Concretely, BML-Trans achieves an average Avg-F1 of 82.5% on CAIL2019 across the labor, divorce, and loan domains, outperforming the strongest baseline by 1.2 percentage points overall and by up to 5.7 percentage points on tail-label Macro-F1, at only 14% additional training cost. Ablation confirms a cascade dependency among the three components, establishing that the gains are structural rather than incidental to threshold selection or initialization. Full article

This study develops a hybrid framework integrating ensemble learning with explainable artificial intelligence to address the methodological challenge of balancing predictive accuracy and interpretability in credit risk model comparison. Using the German Credit Dataset, we implemented a comprehensive preprocessing pipeline, including feature encoding, scaling, and SMOTE for class imbalance handling. Four base models, logistic regression, Random Forest, XGBoost, and Multilayer Perceptron, were combined through a Stacked Ensemble with a logistic regression meta learner. The ensemble demonstrated strong performance, achieving an AUC of 0.761, precision of 0.783, recall of 0.806, and an F1 score of 0.794, which represented the highest scores among all models tested. Notably, Random Forest (AUC = 0.749) surpassed XGBoost (AUC = 0.733), challenging conventional algorithmic hierarchies. SHAP analysis provided transparent global and local interpretability, identifying Current Account status (SHAP = 0.153), Loan Duration (0.064), and Savings Account (0.063) as dominant predictor variables. Class-imbalance handling and threshold optimisation enhanced practical utility by reducing false positives from 39 to 16, thereby aligning with financial risk priorities. The framework provides a reproducible methodological pipeline for systematically comparing credit scoring approaches, demonstrating how predictive performance can be evaluated alongside interpretability considerations within a benchmark dataset context. Full article

Economic crises significantly restrict corporate access to external financing, and regional differences in recovery capacity deserve close attention. This study examines the financial structure and debt of large enterprises in the Visegrád Four (V4) countries (Hungary, Czechia, Poland, Slovakia), focusing on firms with annual revenues above €10 million. Using data from 2021 to 2023, the analysis explores the relationship between corporate debt—including total debt and loan volumes—and regional economic characteristics at the NUTS 3 level. Financial indicators are assessed in comparison with regional productivity data and a sector-specific specialization index sourced from Eurostat. The analysis targets the post-COVID-19 recovery period, which significantly influenced corporate financial behavior. The results indicate that corporate debt increased sharply at the onset of the COVID-19 pandemic and subsequently declined, while remaining strongly concentrated in capital regions. Higher firm concentration and employment scale are associated with greater regional indebtedness, whereas stronger productive capacity is linked to lower reliance on external debt outside metropolitan cores. Overall, the findings highlight pronounced structural and regional heterogeneity, illustrating how spatial concentration and underlying regional characteristics shape corporate debt dynamics during periods of economic stress. Full article

Rapid urbanization and infrastructure expansion in Ghana have intensified demand for concrete, yet reliable, context-specific forecasting tools to support long-term infrastructure planning and resource management remain limited. Existing demand models are largely developed for advanced economies or focused on cement production rather than final concrete consumption, limiting the applicability to rapid urbanizing developing countries. This study addresses this gap by developing an integrated forecasting framework to quantify and project concrete demand in Ghana. Using time-series data spanning 2000–2025, the study employs a modelling approach that combines the Autoregressive Distributed Lag (ARDL) model and Error Correction Model (ECM) to examine both short- and long-run relationships between concrete consumption and key macroeconomic indicators, including GDP, population, GDP growth, concrete prices, housing loan interest rates, lending rates, and exchange rates. Forecast results for 2025–2030 indicated a sustained upward trend in concrete consumption, increasing from 39,278.52 m³ in 2026 to 99,430.53 m³ in 2030, with an average annual growth rate of 26.3% and a mean projected demand of 67,730.83 m³. Model evaluation metrics demonstrated high predictive accuracy, confirming the robustness of the proposed framework. The study contributes to the literature on construction demand forecasting by providing a context-specific, empirically validated model of concrete consumption in a developing economy. The findings offer actionable insights for policymakers, urban planners, and construction managers, underscoring the need to proactively scale local production capacity, strengthen supply-chain logistics, and promote sustainable material sourcing to support infrastructure development. Full article

Financial fraud has expanded rapidly with the growth of the digital economy, evolving from conventional transactional misconduct to more complex and data-intensive forms. Traditional rule-based detection methods are increasingly inadequate for addressing the scale, heterogeneity, and dynamic behavior of modern fraud. In this context, artificial intelligence (AI) has become a core tool in financial fraud detection research. This review systematically surveys AI-based financial fraud detection studies published between 2015 and 2025. It summarizes representative machine learning and deep learning approaches, including tree-based models, neural networks, and graph-based methods, and examines their applications in major fraud scenarios such as credit card fraud, loan fraud, and anti-money laundering. In addition, emerging research on cryptocurrency- and blockchain-related fraud is reviewed, highlighting the distinct challenges posed by decentralized transaction environments. Through a comparative analysis of methods, datasets, and evaluation practices, this review identifies persistent issues in the literature, including severe class imbalance, concept drift, limited access to labeled data, and trade-offs between detection performance and interpretability. Based on these findings, the paper discusses practical considerations for applied fraud detection systems and outlines future research directions from a data-centric and application-oriented perspective. This review aims to provide a structured reference for researchers and practitioners working on real-world financial fraud detection problems. Full article

Amid global population growth and the pressures of climate change, traditional factors in driving grain production are increasingly inadequate. In this context, this study utilizes panel data from 31 provinces in China (2001–2022) to examine the impact of digitalization (DG), technological innovation (TI), and financial innovation (FI) on grain production (GP). Specifically, the study constructs a comprehensive evaluation framework where DG is measured by mobile communication penetration, TI by agricultural R&D expenditure, and FI by the loan-to-GDP ratio of financial institutions. Employing a fixed effects model, the empirical results demonstrate that DG, TI, and FI yield significant positive returns on GP, with elasticity coefficients of 0.262, 10.675 and 6.589, respectively. Furthermore, synergistic effects are observed, where the interaction between DG and TI (coefficient 0.0402 × 10⁻³) and between DG and FI (coefficient 0.221 × 10⁻²) significantly amplify production capacity. Channel analysis reveals that DG and TI promote GP by fostering agricultural industrial agglomeration, TI and FI by advancing agricultural mechanization, and DG and FI by expanding the agricultural insurance scale. Heterogeneity analysis indicates that these factors significantly boost GP in major grain-producing regions but are insignificant in grain production–marketing balanced regions. In major grain-consuming regions, DG has an adverse effect, TI has no significant impact, and FI has a positive effect. Additionally, in non-humid regions, the positive impacts of DG, TI, and FI on GP are greater. Full article

This study investigates credit unions’ expansion into non-core lending and its association with risk and financial resilience. Using US credit union call report data from 1994 to 2024, we measure the share of purchased loans, lease receivables, and loans held for sale in non-core lending. We document robust conditional, within-credit-union associations that point to a clear risk trade-off. Credit unions with higher non-core exposure grow faster in terms of loans and membership but exhibit weaker financial buffers, including lower net worth ratios and weaker economic solvency, alongside higher delinquency. Decomposition tests indicate that loans held for sale are most strongly associated with adverse buffer and asset quality patterns, while purchased loans and lease receivables display smaller and less uniform relationships. Scale interactions suggest that these associations are generally weaker for larger institutions for both membership and assets. Post-COVID estimates indicate that the baseline relationships are broadly stable, while the growth link is becoming stronger. Full article

Credit risk refers to the possibility that a borrower fails to meet contractual repayment obligations, posing potential losses to lenders. This study aims to enhance post-loan default prediction in credit risk management by constructing a time-series modeling framework based on repayment behavior data, enabling the capture of repayment risks that emerge after loan issuance. To achieve this objective, a Residual Enhanced Encoder Bidirectional Long Short-Term Memory (ResE-BiLSTM) model is proposed, in which the attention mechanism is responsible for discovering long-range correlations, while the residual connections ensure the preservation of distant information. This design mitigates the tendency of conventional recurrent architectures to overemphasize recent inputs while underrepresenting distant temporal information in long-term dependency modeling. Using the real-world large-scale Freddie Mac Single-Family Loan-Level Dataset, the model is evaluated on 44 independent cohorts and compared with five baseline models, including Long Short-Term Memory (LSTM), Bidirectional LSTM (BiLSTM), Gated Recurrent Unit (GRU), Convolutional Neural Network (CNN), and Recurrent Neural Network (RNN) across multiple evaluation metrics. The experimental results demonstrate that ResE-BiLSTM achieves superior performance on key indicators such as F1 and AUC, with average values of 0.92 and 0.97, respectively, and demonstrates robust performance across different feature window lengths and resampling settings. Ablation experiments and SHapley Additive exPlanations (SHAP)-based interpretability analyses further reveal that the model captures non-monotonic temporal importance patterns across key financial features. This study advances time-series–based anomaly detection for credit risk prediction by integrating global and local temporal learning. The findings offer practical value for financial institutions and risk management practitioners, while also providing methodological insights and a transferable modeling paradigm for future research on credit risk assessment. Full article

Green hydrogen is gaining global attention as a zero-carbon energy carrier with the potential to drive sustainable energy transitions, particularly in regions facing rising fossil fuel costs and resource depletion. In sub-Saharan Africa, financing mechanisms and structured off-take agreements are critical to attracting investment across the green hydrogen value chain, from advisory and pilot stages to full-scale deployment. While substantial funding is required to support a green economic transition, success will depend on the effective mobilization of capital through smart public policies and innovative financial instruments. This review evaluates financing mechanisms relevant to sub-Saharan Africa, including green bonds, public–private partnerships, foreign direct investment, venture capital, grants and loans, multilateral and bilateral funding, and government subsidies. Despite their potential, current capital flows remain insufficient and must be significantly scaled up to meet green energy transition targets. This study employs a mixed-methods approach, drawing on primary data from utility firms under the H₂Atlas-Africa project and secondary data from international organizations and the peer-reviewed literature. The analysis identifies that transitioning toward Net-Zero emissions economies through hydrogen development in sub-Saharan Africa presents both significant opportunities and measurable risks. Specifically, the results indicate an estimated investment risk factor of 35%, reflecting potential challenges such as financing, infrastructure, and policy readiness. Nevertheless, the findings underscore that green hydrogen is a viable alternative to fossil fuels in sub-Saharan Africa, particularly if supported by targeted financing strategies and robust policy frameworks. This study offers practical insights for policymakers, financial institutions, and development partners seeking to structure bankable projects and accelerate green hydrogen adoption across the region. Full article

Population aging poses significant challenges to the sustainability of pension systems. This study presents an integrated methodological approach that uniquely combines actuarial life-cycle modeling with agent-based simulation to assess the potential of Reverse Mortgage Loans (RMLs) as a dual lever for enhancing retiree welfare and supporting pension system resilience under demographic and financial uncertainty. We explore Reverse Mortgage Loans (RMLs) as a potential financial instrument to support retirees while alleviating pressure on public pensions. Unlike prior research that treats individual decisions or policy outcomes in isolation, our hybrid model explicitly captures feedback loops between household-level behavior and system-wide financial stability. To test our hypothesis that RMLs can improve individual consumption outcomes and bolster systemic solvency, we develop a hybrid model combining actuarial techniques and agent-based simulations, incorporating stochastic housing prices, longevity risk, regulatory capital requirements, and demographic shifts. This dual-framework enables a structured investigation of how micro-level financial decisions propagate through market dynamics, influencing solvency, pricing, and adoption trends. Our central hypothesis is that reverse mortgages, when actuarially calibrated and macroprudentially regulated, enhance individual financial well-being while preserving long-run solvency at the system level. Simulation results indicate that RMLs can improve consumption smoothing, raise expected utility for retirees, and contribute to long-term fiscal sustainability. Moreover, we introduce a dynamic regulatory mechanism that adjusts capital buffers based on evolving market and demographic conditions, enhancing system resilience. Our simulation design supports multi-scenario testing of financial robustness and policy outcomes, providing a transparent tool for stress-testing RML adoption at scale. These findings suggest that, when well-regulated, RMLs can serve as a viable supplement to traditional retirement financing. Rather than offering prescriptive guidance, this framework provides insights to policymakers, financial institutions, and regulators seeking to integrate RMLs into broader pension strategies. Full article

Weeds cause significant yield and economic losses by competing with crops and increasing production costs. Compounding these challenges are labor shortages, herbicide resistance, and environmental pollution, making weed management increasingly difficult. In response, precision weed control (PWC) technologies, such as robots and unmanned aerial vehicles (UAVs), have emerged as innovative solutions. These tools offer farmers high precision (±1 cm spatial accuracy), enabling efficient and sustainable weed management. Herbicide spraying robots, mechanical weeding robots, and laser-based weeders are deployed on large-scale farms in developed countries. Similarly, UAVs are gaining popularity in many countries, particularly in Asia, for weed monitoring and herbicide application. Despite advancements in robotic and UAV weed control, their large-scale adoption remains limited. The reasons for this slow uptake and the barriers to widespread implementation are not fully understood. To address this knowledge gap, our review analyzes 155 articles and provides a comprehensive understanding of PWC challenges and needed interventions for scaling. This review revealed that AI-driven weed mapping in robots and UAVs struggles with data (quality, diversity, bias) and technical (computation, deployment, cost) barriers. Improved data (collection, processing, synthesis, bias mitigation) and efficient, affordable technology (edge/hybrid computing, lightweight algorithms, centralized computing resources, energy-efficient hardware) are required to improve AI-driven weed mapping adoption. Specifically, robotic weed control adoption is hindered by challenges in weed recognition, navigation complexity, limited battery life, data management (connectivity), fragmented farms, high costs, and limited digital literacy. Scaling requires advancements in weed detection and energy efficiency, development of affordable robots with shared service models, enhanced farmer training, improved rural connectivity, and precise engineering solutions. Similarly, UAV adoption in agriculture faces hurdles such as regulations (permits), limited payload and battery life, weather dependency, spray drift, sensor accuracy, lack of skilled operators, high initial and operational costs, and absence of standardized protocol. Scaling requires financing (subsidies, loans), favorable regulations (streamlined permits, online training), infrastructure development (service providers, hiring centers), technological innovation (interchangeable sensors, multipurpose UAVs), and capacity building (farmer training programs, awareness initiatives). Full article

Pakistan’s banking sector faces a critical juncture as rising competition intersects with uneven efficiency, jeopardizing financial stability. This study employs a two-stage empirical framework: (1) evaluating cost-efficiency (CE) evolution via a novel stochastic frontier analysis (SFA) framework incorporating desirable and undesirable outputs (e.g., nonperforming loans) and (2) assessing competition’s impact using a novel multi-product Lerner index across loan, deposit, and asset markets, analyzed via a two-step dynamic panel data system generalized method of moments. The first stage reveals an average CE of 81%, with significant ownership-based disparities. The second stage shows that market power enhances CE overall, supporting the banking-specificity hypothesis, suggesting that regulators balance competition with operational scale benefits. However, market power exhibits duality such as elevating CE in high-efficiency quartile banks but reducing it in low-efficiency quartile ones, confirming the efficient structure hypothesis. This highlights the need for policies promoting efficiency-driven consolidation and addressing structural bottlenecks in underperforming banks. Bank-specific and macroeconomic factors also significantly influence CE. The findings offer a policy roadmap to cultivate a competitive, efficient banking ecosystem, fostering sustainable economic growth. Full article

In the context of the implementation of the rural revitalization strategy in China, returning rural migrant workers are bound to have a certain impact on the rural economy, and land is a very important factor in the agricultural economy. Using data from the 2018 China Labor Dynamic Survey (CLDS), this study examines how migrant workers’ return behaviors influence farmland transfer-in. To address potential endogeneity, the analysis employs the Probit model, instrumental variable methods, and propensity score matching. The findings reveal that returning migrant workers significantly promote farmland transfer-in. Households with returning migrant workers exhibit stronger demands for land transfer-in and tend to operate farmland on a larger scale. Furthermore, returning migrant workers drive farmland expansion through mechanization labor substitution, enhanced access to agricultural loans, and reduced non-farm participation. Additionally, returning migrant workers who are highly educated and younger play a particularly influential role, underscoring the heterogeneous impacts across different migrant groups. This study provides empirical evidence for rural revitalization policies in China by systematically analyzing the effect of returning migrant workers in promoting land transfer-in and the path of influence on farmland scale. Full article

Traditional centralized energy grids struggle to meet urban areas’ increasingly complex energy demands, necessitating the development of more sustainable and resilient energy solutions. Smart microgrids offer a decentralized approach that enhances energy efficiency, facilitates the integration of renewable energy sources, and improves urban resilience. This study follows a systematic review approach, analyzing the literature published in peer-reviewed journals, conference proceedings, and industry reports between 2011 and 2025. The research draws from academic publications of energy institutions alongside regulatory reports, examining actual smart microgrid deployments in San Diego, Barcelona, and Seoul. Additionally, this article provides real-world case studies from New York and London, showcasing successful and unsuccessful smart microgrid deployments. The Brooklyn Microgrid in New York demonstrates peer-to-peer energy trading, while London faces regulations and funding challenges in its decentralized energy systems. The paper also explores economic and policy frameworks such as public–private partnerships (PPPs), localized energy markets, and standardized regulatory models to enable microgrid adoption at scale. While PPPs provide financial and infrastructural support for microgrid deployment, they also introduce stakeholder alignment and regulatory compliance complexities. Countries like Germany and India have successfully used PPPs for smart microgrid development, leveraging low-interest loans, government incentives, and regulatory mechanisms to encourage innovation and adoption of smart microgrid technologies. In addition, the review examines new trends like the utilization of AI and quantum computing to optimize energy, peer-to-peer energy trading, and climate resilient design before outlining a future research agenda focused on cybersecurity, decarbonization, and the inclusion of new technology. Contributions include the development of a modular and scalable microgrid framework, innovative hybrid storage systems, and a performance-based policy model suited to the urban environment. These contributions help to fill the gap between what is possible today and what is needed for future sustainable urban energy systems and create the foundation for resilient cities of the next century. Full article