Search Results (4,215)

Viticulture is a vital sector of agriculture and economy exhibiting susceptibility to climate change, particularly in the Mediterranean regions. The present investigation examines the climatic suitability for vineyards development in Greece by exploiting geomorphological and bioclimatic data for the reference climatic period 1970–2000. The data is sourced from the ERA5-Land dataset and analyzed with R. The objective is to create a specific crop suitability map based on a simple, transparent model implemented through coding. This map identifies the climatically suitable areas for grapevine cultivation during the reference period. Results demonstrate that the model is highly adaptable, as both variable thresholds and areas of interest can be modified, while incorporating future climate scenarios can be performed, allowing for dynamic reconfiguration. According to the mapped climatic suitability, 55.1% of Greece is rated 3.5–4.0, and 12.9% is rated 4.0–4.5. The total suitability over Greece is calculated with a score of 3.5–4.0 for the 50.9% of total area, and for a score of 4.0–4.5, the covered area is 12.9%. Considering the Corine Land Cover classification as the reference land cover dataset, the false-negative areas (the model indicates that an area with vines is not suitable) are only 1.5% of the areas defined as viticultural. By providing clear and accurate spatial information, the model supports informed decision-making and the development of adaptation strategies, enhancing, therefore, the resilience and sustainability of viticulture in the context of climate change. Full article

Detecting minute defects, such as spurs on the surface of a Printed Circuit Board (PCB), is extremely challenging due to their small size (average size < 20 pixels), sparse features, and high dependence on circuit topology context. The original Mosaic data augmentation method faces significant challenges with semantic adaptability when dealing with such tasks. Its unrestricted random cropping mechanism easily disrupts the topological structure of minute defects attached to the circuits, leading to the loss of key features. Moreover, a splicing strategy without domain constraints struggles to simulate real texture interference in industrial settings, making it difficult for the model to adapt to the complex and variable industrial inspection environment. To address these issues, this paper proposes a Context-aware and Domain-consistent Mosaic (CD-Mosaic) augmentation algorithm. This algorithm abandons pure randomness and constructs an adaptive augmentation framework that synergizes feature fidelity, geometric generalization, and texture perturbation. Geometrically, an intelligent sampling and dynamic integrity verification mechanism, driven by “utilization-centrality”, is designed to establish a controlled sample quality distribution. This prioritizes the preservation of the topological semantics of dominant samples to guide feature convergence. Meanwhile, an appropriate number of edge-truncated samples are strategically retained as geometric hard examples to enhance the model’s robustness against local occlusion. For texture, a dual-granularity visual perturbation strategy is proposed. Using a homologous texture library, a hard mask is generated in the background area to simulate foreign object interference, and a local transparency soft mask is applied in the defect area to simulate low signal-to-noise ratio imaging. This strategy synthesizes visual hard examples while maintaining photometric consistency. Experiments on an industrial-grade PCB dataset containing 2331 images demonstrate that the YOLOv11m model equipped with CD-Mosaic achieves a significant performance improvement. Compared with the native Mosaic baseline, the core metrics mAP@0.5 and Recall reach 0.923 and 86.1%, respectively, with a net increase of 8.3% and 8.8%; mAP@0.5:0.95 and AP_small, which characterize high-precision localization and small target detection capabilities, are improved to 0.529 (+3.0%) and 0.534 (+3.3%), respectively; the comprehensive metric F1-score jumps to 0.903 (+6.2%). The experiments prove that this method effectively solves the problem of missed detections of industrial minute defects by balancing sample quality and detection difficulty. Moreover, the inference speed of 84.9 FPS fully meets the requirements of industrial real-time detection. Full article

Background: Heart rate variability (HRV) is a valuable indicator for assessing vascular health, but keeping clinical decision support systems (CDSSs) aligned with the rapidly evolving literature remains challenging. This study aimed to develop an LLM-assisted literature synthesis framework to support transparent HRV-based risk stratification, enabling systematic extraction and organization of HRV evidence from published studies. Methods: An LLM-driven framework was developed to extract HRV parameters from 140 medical abstracts. The system simulated step-by-step human reasoning to identify key HRV indicators and group patient data using predefined statistical thresholds derived from the literature. System performance was evaluated using ECG-derived HRV features as a feasibility evaluation of literature-guided HRV classification. Results: The proposed framework demonstrated an accuracy of 86% in literature-guided HRV classification, with a sensitivity of 81% and a specificity of 87%. Compared with traditional machine learning approaches, the LLM-assisted system provided transparent, literature-grounded reasoning and could be readily updated as new studies became available. Conclusions: Large language models can support evidence-guided parameter selection and feasibility-level HRV-based risk stratification, rather than serving as predictive classifiers. This approach reduces manual effort, enhances transparency, and addresses common “black box” concerns associated with AI-assisted CDSS development in clinical practice. Full article

As global climate and sustainable challenges gain more attention, green finance has emerged as a significant focus of worldwide financial reform, with green bonds serving as a key indicator. Retail investors, as an important part of the financial market, have a significant impact on the development of green finance through their investment willingness. This study aims to explore the influencing factors and mechanisms on the investment willingness of China retail investors towards green bonds. Based on empirical analysis of data from 2219 valid respondents in China, carried out using the SEM method, the results suggest that perceived usefulness (PU), investment literacy (IL), and information transparency (IT) all positively influence retail investors’ willingness to invest in green bonds. Additionally, PU, IL, and IT contribute to fostering an open attitude toward change (OATC) among retail investors, which, in turn, significantly promotes their investment willingness. This study also identifies the mediation effect of OATC. The findings provide both theoretical and practical insights to promote the development of green finance, enhance market activity, and support policy frameworks. Full article

This paper presents an MAUT-based decision-support framework, developed within the NAVGREEN project, to enable the evaluation of alternative fuels and technologies in shipping decarbonization pathways toward zero-emission targets. The framework integrates stakeholder-derived weights elicited through the Analytic Hierarchy Process (AHP) and systematically evaluates alternatives across five criteria: cost, technological maturity, safety and regulatory compatibility, carbon footprint, and social acceptability. Alternatives are mapped into a common utility space through criterion-specific utility functions and aggregated into a composite utility score, enabling transparent and reproducible comparison of single and combined solutions. To strengthen applicability beyond a single illustrative application, the study incorporates a structured scenario and sensitivity analyses (policy stringency, infrastructure constraints, conservative regulatory environments, and weight and parameter perturbations) to assess ranking stability under plausible future conditions. A case study on an Ultramax bulk carrier is used solely to demonstrate the operability and workflow of the method, rather than to empirically validate technology choices across all ship types. Optional AI-assisted elicitation may be used as a supporting aid to harmonize indicative inputs when data are incomplete; however, validation of AI-generated estimates is outside the scope of the present study and is identified as future work. Full article

This study develops a reproducible regional screening workflow to assess geothermal potential in the Cambrian reservoir system of Western Lithuania under conditions of sparse and heterogeneous legacy subsurface data. The approach integrates data compilation, cleaning, and harmonization from archival well materials, ordinary kriging spatialization of key reservoir properties with uncertainty multipliers, standardized doublet simulations to derive comparative thermal performance indicators, and a neural network surrogate to accelerate regional inference. The workflow integrates 12 compiled reservoir control points into a gridded regional representation (25 × 30 cells; ~6750 km²) and evaluates uncertainty through low, mid and high scenarios (±10%). Physics-based simulations were executed for 303 representative grid locations per scenario, yielding cumulative extracted-energy indicators on the order of 10⁵–10⁷ MWh across cases (reported as comparative indicators). The neural network surrogate reproduced simulation outputs with a high predictive agreement (test R² = 0.996; cross-validation mean R² ≈ 0.99), enabling swift prediction across the remaining grid cells after training. Relative potential maps highlight spatially coherent zones of higher prospectivity and provide a transparent basis for prioritizing follow-up investigations and data acquisition. The proposed framework is modular and can be refined as improved geological constraints, thermophysical properties, and operational assumptions become available. Full article

Modern Autonomous Vehicles generate large volumes of heterogeneous in-vehicle data, making cybersecurity a critical challenge as adversarial attacks become increasingly adaptive, stealthy, and multi-protocol. Traditional intrusion detection systems often fail under these conditions because of their limited contextual understanding, poor robustness to distribution shifts, and insufficient regulatory transparency. This study introduces LLM-Guardian, a hierarchical intrusion detection framework with decision-making mechanisms that integrates Large Language Models (LLMs) with classical statistical detection theory, optimal transport drift analysis, graph neural networks, and formal uncertainty quantification. LLM-Guardian uses semantic anomaly scoring, conformal prediction for distribution-free confidence calibration, adaptive cumulative sum (CUSUM) sequential testing for low-latency detection, and topology-aware GNN reasoning designed to identify coordinated attacks across CAN, Ethernet, and V2X interfaces. In this work, the framework is empirically evaluated on four heterogeneous CAN-bus datasets, while the Ethernet and V2X components are instantiated at the architectural level and left as directions for future multi-protocol experimentation. Full article

In the context of China’s “dual-carbon” strategy, the Beijing 2022 Winter Olympics provides a critical case for examining whether carbon-neutral commitments can be translated into measurable and lasting environmental outcomes through a closed-loop governance mechanism. This study develops an integrated analytical framework linking institutional design, policy tools, monitoring–reporting–verification (MRV), and environmental legacy, and evaluates full life-cycle carbon-neutral governance and post-event environmental performance using officially verified carbon accounting materials, governmental disclosures, and publicly available statistical data from 2016–2022. We synthesize the emission structure across preparation and Games-time phases, examine key mitigation and offset portfolios, and assess multi-dimensional environmental indicators in Beijing and Zhangjiakou, including atmospheric quality, energy structure transition, ecological restoration, and low-carbon transport systems. The results suggest that an MRV-centered governance chain strengthened accounting transparency and compliance-oriented implementation, while environmental indicators in the competition zones exhibited sustained improvement over the study period. To reduce over-attribution under concurrent national clean-air policies and macro-level environmental governance trends, we benchmarked host-zone indicators against external reference statistics and interpret the observed improvements as an “acceleration effect” under bounded inference rather than a strict net causal contribution. The findings highlight the importance of hotspot-oriented asset-chain governance (transport infrastructure and venue construction), robust MRV disclosure, and quality-controlled offsets in shaping credible environmental legacies, and provide policy implications for future mega-events seeking to balance carbon neutrality with long-term regional sustainability. Full article

Romania is highly exposed to seismic risk, with significant implications for residential earthquake insurance and risk-transfer mechanisms, due to the Vrancea intermediate-depth seismic source and a vulnerable building stock. This paper presents a harmonised seismic exposure model for the Romanian residential sector, developed to support probabilistic seismic risk assessment and catastrophe modelling for (re)insurance applications. The model integrates official data from the 2021 Population and Housing Census with the nationally adopted RTC-10 structural typology, height classification, seismic code level, and standardised reconstruction cost indicators. The results indicate that nearly 70% of residential dwellings were constructed before 1990 under pre-code or low- to moderate-code seismic design provisions. Although individual houses dominate the dwelling stock, multi-family apartment buildings concentrate approximately 40% of the total residential replacement cost, particularly in urban areas. The total replacement cost of the residential building stock is estimated at approximately EUR 709 billion, exceeding values derived from global exposure models. Comparison with existing insurance coverage highlights a substantial protection gap between potential seismic losses and insured values. The proposed exposure model provides a transparent, nationally calibrated basis for seismic loss estimation, portfolio accumulation analysis, and evidence-based risk management in both engineering and (re)insurance contexts. Full article

Modern malware presents significant challenges to traditional detection methods, often leveraging fileless techniques, in-memory execution, and process injection to evade antivirus and signature-based systems. To address these challenges, alert-driven memory forensics has emerged as a critical capability for uncovering stealthy, persistent, and zero-day threats. This study presents a two-stage host-based malware detection framework, that integrates memory forensics, explainable machine learning, and ensemble classification, designed as a post-alert asynchronous SOC workflow balancing forensic depth and operational efficiency. Utilizing the MemMal-D2024 dataset—comprising rich memory forensic artifacts from Windows systems infected with malware samples whose creation metadata spans 2006–2021—the system performs malware detection, using features extracted from volatile memory. In the first stage, an Attentive and Interpretable Learning for structured Tabular data (TabNet) model is used for binary classification (benign vs. malware), leveraging its sequential attention mechanism and built-in explainability. In the second stage, a Voting Classifier ensemble, composed of Light Gradient Boosting Machine (LGBM), eXtreme Gradient Boosting (XGB), and Histogram Gradient Boosting (HGB) models, is used to identify the specific malware family (Trojan, Ransomware, Spyware). To reduce memory dump extraction and analysis time without compromising detection performance, only a curated subset of 24 memory features—operationally selected to reduce acquisition/extraction time and validated via redundancy inspection, model explainability (SHAP/TabNet), and training data correlation analysis —was used during training and runtime, identifying the best trade-off between memory analysis and detection accuracy. The pipeline, which is triggered from host-based Wazuh Security Information and Event Management (SIEM) alerts, achieved 99.97% accuracy in binary detection and 70.17% multiclass accuracy, resulting in an overall performance of 87.02%, including both global and local explainability, ensuring operational transparency and forensic interpretability. This approach provides an efficient and interpretable detection solution used in combination with conventional security tools as an extra layer of defense suitable for modern threat landscapes. Full article

Rapid urbanization is driving sharp growth in construction and demolition waste (CDW), making recycling facility selection and transport planning critical for cost-effective and sustainable urban waste management. This paper presents an end-to-end, simulation-driven decision-support framework that jointly optimizes facility selection and operational waste transportation policies under uncertainty, and systematically benchmarks competing solutions using Data Envelopment Analysis (DEA). The proposed approach embeds a metaheuristic optimization engine within a Monte Carlo simulation environment to evaluate facility configurations and dispatch–allocation decisions under stochastic waste generation and operating conditions, using sample-average performance to ensure fair and consistent comparison across scenarios. Results from the Wuhan metropolitan case study show that coordinating dispatch intensity with contracted facility capacity significantly reduces total cost and unmoved waste while stabilizing performance across stochastic realizations; DEA then provides transparent efficiency-frontier ranking across economic, operational, and environmental indicators without requiring pre-specified weights. These findings demonstrate that dispatch–capacity alignment is a dominant lever for robust and sustainable CDW logistics, and that DEA-based benchmarking enhances decision transparency when multiple near-optimal solutions coexist. Full article

Hybrid renewable energy systems are increasingly important for enabling sustainable and resilient energy supply in rural smart communities, yet existing tools often lack the ability to integrate environmental variability, multi-technology interactions, and economic–environmental assessment in a unified framework. This study presents Hybrid Smart Micro Energy Community (HySMEC), a novel modeling approach that combines high-resolution meteorological data, technology-specific generation models, detailed demand characterization, and financial analysis to evaluate hybrid configurations of hydropower, solar PV, wind, battery storage, and grid interaction. Hourly simulations capture seasonal dynamics and system behavior under realistic technical efficiencies, investment costs, and emission factors, enabling a transparent assessment of energy flows, self-consumption, and grid dependence. The results show that hybrid systems can achieve competitive economic performance, low Levelized Costs of Energy, and significant CO₂ emission reductions across diverse rural community profiles, even when space or demand constraints are present. The analysis confirms the technical feasibility and environmental benefits of integrating multiple renewable sources with storage, highlighting the importance of self-consumption ratios in improving system profitability. Overall, HySMEC provides a robust and scalable tool to support data-driven design and optimization of distributed energy systems, offering valuable insights for researchers, planners, and decision-makers involved in sustainable rural energy development. Full article

Background: This study aimed to create and evaluate explainable machine learning models for forecasting postoperative stone-free status following percutaneous nephrolithotomy (PNL) utilizing a substantial clinical cohort. Methods: This retrospective single-center analysis encompassed 2144 adult patients who received PNL from 2010 to 2024. We employed clinical, radiographic, stone-related, and surgical data to train four supervised machine learning models: Extreme Gradient Boosting (XGBoost), Random Forest, Light Gradient Boosting Machine (LightGBM), and Adaptive Boosting (AdaBoost). We used the Synthetic Minority Oversampling Technique exclusively on the training set to fix the class imbalance. We assessed the model’s accuracy, precision, recall, F1-score, and area under the receiver operating characteristic curve (ROC–AUC) to see how well it worked. SHapley Additive exPlanations (SHAP) were used to measure explainability. Results: The total stone-free rate was 84.8%. XGBoost had the best predictive performance of the models tested, with an accuracy of 0.916 and a ROC–AUC of 0.975. LightGBM was close behind. Random Forest and AdaBoost had relatively inferior performance. SHAP analysis identified anatomical anomalies as demonstrated the strongest association with stone-free outcomes. The size of the access sheath and the number of stones were next. Other parameters that were identified by SHAP as important contributors to model predictions were the placement of the stone, Guy’s Stone Score, the length of the operation, and the density of the stone. These feature associations demonstrated clinical coherence with established knowledge in surgical practice. Conclusions: Explainable machine learning algorithms, especially XGBoost, can accurately predict stone-free outcomes following PNL in a way that makes sense to doctors. The incorporation of SHAP improves transparency and facilitates the prospective application of these models as decision-support instruments in personalized surgical planning. Full article

Climate change increases uncertainty in agricultural production and rural livelihoods, encouraging farms to pursue diversification strategies to buffer climate-related risks. Concurrently, the growing use of digital and AI-based climate decision-support tools raises questions about how the transparency of such information shapes farm-level adaptation. This study examines the relationships among AI transparency, climate awareness, decision confidence, agritourism diversification intention, and perceived farm resilience within a perception-based analytical framework in climate-sensitive rural systems. Data were collected through in-person fieldwork conducted throughout 2025 among agritourism-oriented farm operators in two Serbian rural clusters: a Western mountain agritourism belt and an Eastern/Southeastern dry-stress zone. Using structural equation modeling, the analysis reveals a consistent pattern of positive associations across all modeled relationships. Higher perceived transparency of AI-based climate information is associated with stronger climate awareness, greater decision confidence, a higher intention to diversify toward agritourism, and greater perceived farm resilience. Perceived farm resilience was most strongly related to agritourism diversification intention, underscoring diversification as a key perceived adaptive pathway under climate stress. The findings highlight AI transparency as a critical informational precondition for adaptive decision-making and resilience building as evaluated by farm operators, with implications for farmer-centric digital tools and rural climate adaptation policy in comparable climate-sensitive agricultural contexts. Full article

Embodied-carbon accounting is increasingly required at the early design stage to guide material and construction choices during design iterations. However, many life-cycle assessment (LCA) workflows and centralized building information modeling (BIM)–LCA plugins still rely on fragmented data, non-transparent mapping rules, and limited cross-project reuse, which slows rapid iteration. This study develops an open and traceable embodied-carbon assessment workflow driven by BIM object geometry and semantic attributes and demonstrates it through a single case study, enabling automated accounting for the A1–A3 stages from model input to result reporting. The framework is implemented as a Revit add-in prototype connected to an open-data platform. It uses assemblies as standardized assessment units, applies configurable rule-based mapping, and performs unit normalization to link model quantities with carbon factors. A single three-story brick–concrete residential building in Wuhan with an LoD 300 model is used as the sole validation case to demonstrate workflow feasibility, report coverage, and time metrics. The case yields an A1–A3 embodied-carbon intensity of approximately 333 kgCO₂ e/m², dominated by the structural system. Rule mapping achieves 82% coverage within the defined accounting scope. Compared with manual workflows (290–380 min), first-time accounting is reduced to 83–98 min and further to within 30 min when assemblies and rules are reused. Contribution decomposition shows a concentrated pattern and supports traceability from assemblies to material types. Overall, within the tested scope, the Revit-based prototype provides efficient and verifiable embodied-carbon feedback for early-stage design. Full article