Search Results (2,845)

Credit card fraud detection remains a major challenge due to extreme class imbalance and evolving attack patterns. This paper proposes a practical hybrid pipeline that combines conditional tabular generative adversarial networks (CTGANs) for targeted minority-class synthesis with Light Gradient Boosting Machine (LightGBM) for classification. Inspired by symmetry principles in machine learning, we leverage the adversarial equilibrium of CTGAN to generate realistic fraudulent transactions that maintain distributional symmetry with real fraud patterns, thereby preserving the structural and statistical balance of the original dataset. Synthetic fraud samples are merged with real data to form augmented training sets that restore the symmetry of class representation. We evaluate Simple Recurrent Neural Network (RNN), Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU) classifiers, and a LightGBM model on a public dataset using stratified 5-fold validation and an independent hold-out test set. Models are compared using sensitivity, precision, F-measure(F₁), and area under the precision–recall curve (PR-AUC), which reflects symmetry between detection and false-alarm trade-offs. Results show that CTGAN-based augmentation yields large and consistent gains across architectures. The best-performing configuration, CTGAN + LightGBM, attains sensitivity = 0.986, precision = 0.982, F₁ = 0.984, and PR-AUC = 0.918 on the test data, substantially outperforming non-augmented baselines and recent methods. These findings indicate that conditional synthetic augmentation materially improves the detection of rare fraud modes while preserving low false-alarm rates, demonstrating the value of symmetry-aware data synthesis in classification under imbalance. We discuss generation-quality checks, risk of distributional shift, and deployment considerations. Future work will explore alternative generative models with explicit symmetry constraints and time-aware production evaluation. Full article

Reliable measurement of multiphase flow is fundamental to production evaluation in complex oil and gas wells. However, conventional sensors often suffer from low integration, limited measurement capability, and potential environmental impact. To address these challenges, a photoelectric composite three-phase flow sensor is developed, integrating multiple electrode rings for water holdup and liquid-phase velocity measurement, with dual optical-fiber probes for gas holdup and gas-phase velocity detection. A slip model is further applied to quantify the dependence of slip velocity on liquid holdup based on measured phase rates. Experimental results demonstrate high sensitivity to bubble-flow structures, accurate extraction of gas holdup and phase velocities, and stable full-range water holdup calibration from 0% to 100% at 5 V and 15 V with effective temperature and salinity compensation. And compared with existing technologies, the sensor designed in this paper has the advantages of high integration, a simple structure, multiple measurement parameters, and higher water-holding capacity resolution in low-saturation areas, providing more advanced technical means for conventional profile three-phase flow logging. Full article

Lattice-based post-quantum cryptography relies on fast polynomial multiplication. The Number-Theoretic Transform (NTT) is the key operation that enables this acceleration. To provide high throughput and low latency while keeping the area overhead small, hardware implementations of the NTT is essential. This is particularly true for resource-constrained devices. However, existing NTT accelerators either achieve high throughput at the cost of large area overhead or provide compact designs with limited pipelining and low operating frequency. Therefore, this article presents a compact, seven-stage pipelined NTT accelerator architecture for post-quantum cryptography, using the CRYSTALS–Kyber algorithm as a case study. The CRYSTALS–Kyber algorithm is selected due to its NIST standardization, strong security guarantees, and suitability for hardware acceleration. Specifically, a unified three-stage pipelined butterfly unit is designed using a single DSP48E1 block for the required integer multiplication. In contrast, the modular reduction stage is implemented using a four-stage pipelined, lookup-table (LUT)-only Barrett reduction unit. The term “LUT-only” refers strictly to the reduction logic and not to the butterfly multiplication. Furthermore, two dual-port BRAM18 blocks are used in a ping-pong manner to hold intermediate and final coefficients. In addition, a simple finite-state machine controller is implemented, which manages all forward NTT (FNTT) and inverse NTT (INTT) stages. For validation, the proposed design is realized on a Xilinx Artix-7 FPGA. It uses only 503 LUTs, 545 flip-flops, 1 DSP48E1 block, and 2 BRAM18 blocks. The complete FNTT and INTT with final rescaling require 1029 and 1285 clock cycles, respectively. At 200 MHz, these correspond to execution times of 5.14 µs for the FNTT and 6.42 µs for the INTT. Full article

This study investigated the combined influence of hydrothermal treatment and particle size on the techno-functional properties of defatted coconut residue (DCR) to optimize its use as a hydrocolloid fat replacer. A 3 × 2 factorial design evaluated boiling and autoclaving treatments in combination with coarse and fine milling. Fine particle fractions (boiling-fine [BF] and autoclaved-fine [AF]) were identified as optimal, exhibiting peak water-holding capacity (WHC) (10.95 g/g) and oil-holding capacity (4.57 g/g) due to maximized surface area and thermal unblocking of capillary networks. When incorporated into cookies, all DCR formulations qualified as “reduced-fat” (30% reduction) and “high-fiber” (6 g/100 g) products. Crucially, the extreme WHC of fine fractions induced severe water competition within the dough, leading to a direct inverse correlation with quality, characterized by a restricted spread ratio (6.9) and increased hardness (27 N). Furthermore, thermal leaching of Maillard precursors suppressed excessive browning, improving cookie color. While the BF fraction provided the best functional balance, future research should optimize dough moisture to mitigate the impact of high fiber hydration on texture. These findings demonstrate DCR’s potential for agro-food valorization and improved human health. Full article

Land seismic data are always contaminated by surface waves, which demonstrate strong energy, low velocity, and long vibrations. Such noises often mask deep effective reflections, seriously reducing the data’s signal-to-noise ratio while limiting the imaging accuracy of complex deep structures and the efficiency of hydrocarbon reservoir identification. To address this critical technical bottleneck, this paper proposes a surface wave joint reconstruction method based on stationary phase analysis, combining the cross-correlation seismic interferometry method with the convolutional seismic interferometry method. This approach integrates cross-correlation and convolutional seismic interferometry techniques to achieve coordinated reconstruction of surface waves in both shot and receiver domains while introducing adaptive weight factors to optimize the reconstruction process and reduce interference from erroneous data. As a purely data-driven framework, this method does not rely on underground medium velocity models, achieving efficient noise reduction by adaptively removing reconstructed surface waves through multi-channel matched filtering. Application validation with field seismic data from the piedmont regions of western China demonstrates that this method effectively suppresses high-energy surface waves, significantly restores effective signals, improves the signal-to-noise ratio of seismic data, and greatly enhances the clarity of coherent events in stacked profiles. This study provides a reliable technical approach for noise reduction in seismic data under complex near-surface conditions, particularly suitable for hydrocarbon exploration in regions with developed scattering sources such as mountainous areas in western China. It holds significant practical application value and broad dissemination potential for advancing deep hydrocarbon resource exploration and improving the quality of complex structural imaging. Full article

Cooperative task allocation is one of the critical enablers for multi-Autonomous Underwater Vehicle (AUV) missions, but existing approaches often assume reliable communication that rarely holds in real underwater acoustic environments. We study here the performance and robustness of the Consensus-Based Bundle Algorithm (CBBA) for multi-AUV task allocation under realistically degraded underwater communication conditions with dynamically appearing tasks. An integrated simulation framework that incorporates a Dubins-based kinematic model with minimum turning radius constraints, a configurable underwater acoustic communication model (range, delay, packet loss, and bandwidth), and a full implementation of improved CBBA with new features, complemented by 3D trajectory and network-topology visualization. We define five communication regimes, from ideal fully connected networks to severe conditions with short range and high packet loss. Within these regimes, we assess CBBA based on task allocation quality (total bundle value and task completion rate), convergence behavior (iterations and convergence rate), and communication efficiency (message delivery rate, average delay, and network connectivity), with additional metrics on the number of conflicts during dynamic task reallocation. Our simulation results indicate that CBBA maintains performance close to the optimum when the conditions are good and moderate but degrades significantly when connectivity becomes intermittent. We then introduce a local-communication-based conflict resolution strategy in the face of frequent task conflicts under very poor conditions: neighborhood-limited information exchange, negotiation within task areas, and decentralized local decisions. The proposed conflict resolution strategy significantly reduces the occurrence of conflicts and improves task completion under stringent communication constraints. This provides practical design insights for deploying multi-AUV systems under weak underwater acoustic networks. Full article

As a vital ecological security barrier and climate regulator in northwestern China, the spatial patterns and evolving formation mechanisms of ecosystem services within the Qinghai Lake basin hold significant strategic value for ecological conservation and national park development in the region. This study selected land use data during 2000–2020, integrating the equivalent factor method, spatial correlation analysis, and the geodetector approach to systematically investigate the spatial heterogeneity characteristics of ESV in the Qinghai Lake basin and its corresponding driving mechanisms. The results indicate the following: (1) During the period 2000–2020, grassland consistently constituted the primary land cover category within the Qinghai Lake Basin, accounting for over 60% of the total area; water bodies (16.67%) and unused land (16.56%) represented the secondary land use categories. Over this twenty-year period, the total ESV exhibited a slight increasing trend, rising from USD 30.30 × 10⁸ to USD 30.75 × 10⁸, representing a growth of 0.31%. Regulating services constituted the primary component of ESV. The highest contribution to ESV originated from water bodies, with grassland ranking second. (2) ESV displayed a spatial arrangement marked by “high values in the lake center and low values in the surrounding areas” and “higher values in the southeast and lower values in the northwest.” Its spatial correlation exhibits a pronounced positive relationship. The number of units classified as high-high clusters (primarily water bodies at low elevations) and low-low clusters (mainly grasslands and unused land at high elevations) both increased over the study period, indicating a continuous intensification of ESV spatial agglomeration. (3) Results from the geographical detector reveal that both natural and anthropogenic factors collectively drive the spatial variation in ESV, with natural factors exhibiting stronger explanatory capacity. Among these, elevation and temperature are identified as the dominant drivers of ESV spatiotemporal differentiation. The combined effect of two interacting factors surpasses the influence exerted by any single factor in isolation. This research clarifies that the spatial distribution of ESV in the Qinghai Lake Basin, which features “high values in the lake center and low values in the surrounding areas” as well as “higher values in the southeast and lower values in the northwest,” is jointly shaped by the combined control of vertical zonality governed by topographic and climatic factors and the spatial differentiation of human activities. In low-altitude lakeshore zones, ESV rose as a consequence of water body expansion and the enforcement of ecological conservation measures, leading to the emergence of high-value clusters. In contrast, ESV improvement in high-elevation regions remained limited, constrained by fragile natural conditions and minimal human intervention. The insights derived from this research offer a scientific foundation for refining the “one core, four zones, one ring, multiple points” functional zoning framework of the Qinghai Lake National Park, as well as for developing tailored management approaches suited to distinct elevation-based regions. Full article

Some properties on large contractions in metric spaces are proven. In particular, such contractions are proven to be asymptotically regular. In addition, if the metric space is complete, then the sequences that they generate are bounded, Cauchy, and convergent to a unique fixed point. Also, cyclic large contractions are an area of focus. It is proven that, if subsets of the cyclic disposal are nonempty closed and they intersect, all the sequences are bounded and Cauchy, and they converge to a unique fixed point located in the intersection of such subsets if the metric space is complete. If the subsets have a pair-wise empty intersection, then the boundedness of such sequences is proven without the need to assume the boundedness of the subsets in the cyclic disposal. The convergence of the sequences to a unique limit cycle of best proximity points, with one per subset in the cyclic disposal, is proven provided that the metric space is complete and that one of such subsets is boundedly compact with a singleton best proximity set. For that property to hold, it is not assumed that the remaining best proximity points are necessarily singletons. It has also been proven that all the subsequences contained within each of the subsets are Cauchy and they converge to a unique best proximity point, even if the corresponding best proximity sets is not a singleton. Furthermore, the hypothesis that one of the best proximity sets between adjacent subsets is a singleton can be weakened for any particular cyclic large contraction. Later on, eventual perturbations of the cyclic large self-mappings in normed spaces are discussed. If the norm of the perturbation additive operator is small enough, it is proven that the perturbed cyclic self-mapping maintains the property of being a cyclic large contraction associated with the unperturbed nominal cyclic large contraction. The maximum upper-bound of the perturbed operator ensures that such a property is given in an explicit manner. Full article

The persistent expansion in the intensity and scope of human disturbance has become a key driver of global biodiversity loss, affecting wildlife behavior and population stability across multiple dimensions. As a characteristic symbiotic assemblage in the subalpine forest ecosystems of Sichuan, the giant panda (Ailuropoda melanoleuca), red panda (Ailurus fulgens), and forest musk deer (Moschus berezovskii) exhibit significant research value in their responses to human disturbance. However, existing studies lack systematic analysis of multiple disturbances within the same protected area. This study was conducted in the Sichuan Liziping National Nature Reserve, where infrared camera traps were deployed using a kilometer-grid layout. By integrating spatiotemporal pattern analysis and Generalized Additive Models (GAM), we investigated the characteristics of human disturbance and the response strategies of the three species within their habitats. The results show that: (1) A total of seven types of human disturbance were identified in the reserve, with the top three by frequency being cattle disturbance, goat disturbance, and walking disturbance; (2) Temporally, summer and winter were high-occurrence seasons for disturbance, with peaks around 12:00–14:00, while the giant panda exhibited a bimodal diurnal activity pattern (10:00–12:00, 14:00–16:00), the red panda peaked mainly at 8:00–10:00, and the forest musk deer preferred crepuscular and nocturnal activity—all three species displayed activity rhythms that temporally avoided peak disturbance periods; (3) Spatially, giant pandas were sparsely distributed, red pandas showed aggregated distribution, and forest musk deer exhibited a multi-core distribution, with the core distribution areas of each species spatially segregated from high-disturbance zones; (4) GAM analysis revealed that the red panda responded most significantly to disturbance, the giant panda showed marginal significance, and the forest musk deer showed no significant response. This study systematically elucidates the spatiotemporal differences in responses to multiple human disturbances among three sympatric species within the same landscape, providing a scientific basis for the management of human activities, habitat optimization, and synergistic biodiversity conservation in protected areas. It holds practical significance for promoting harmonious coexistence between human and wildlife. Full article

In recent years, numerous researchers have investigated various preventive strategies against respiratory viruses that pose a threat to human health. This study aims to develop a nasal spray formulation based on the natural polysaccharide xanthan gum (XG) and camostat, and to evaluate its dual protective mechanism at the nasal mucosa, the primary entry point for respiratory viral infections. The efficacy of the formulation was assessed through physicochemical characterization, cell-based assays, and animal experiments. Initially, muco-adhesiveness was evaluated by monitoring the drying dispersion area of the test formulation over time on a Petri dish. The combination of XG and camostat exhibited a dispersion area more than ten times larger than that of each component used alone. The antiviral efficacy was demonstrated in both human nasal epithelial cells (HNEc) and an influenza-infected mouse model. The cell-based experiment demonstrated a significant inhibition of viral penetration and replication. Furthermore, suppression of transmembrane protease, serine 2 (TMPRSS2) expression, a key factor in influenza virus entry, was observed in mouse lung tissues. These findings suggest that the Camostat–Polysaccharide Dual-Action Nasal Spray (CPNS), currently under development, holds promise as a non-invasive, first-line barrier to prevent the initial infection and replication of respiratory viruses. Full article

This study examines the economic feasibility of adopting digital technologies for the early detection of Verticillium wilt in olive cultivation in Northern Evia, Greece. A Net Present Value (NPV) framework with different scenarios was employed to derive three adoption thresholds: the minimum effectiveness required to break even, the maximum tolerable cost at a target effectiveness, and the break-even olive-oil price. The results reveal substantial variability across scenarios, reflecting uncertainty in both disease dynamics and market conditions. Key determinants of feasibility include detection effectiveness, adoption costs, olive oil prices, and disease incidence. Larger holdings consistently face more favourable thresholds due to economies of scale, while smaller farms remain constrained unless collective actions or policy support reduces costs. The preliminary evidence indicates that early detection technologies can strengthen the resilience of olive farms, especially in high-incidence areas, though feasibility remains highly sensitive to costs, prices, and pathogen pressure. Finally, the findings underscore the need for targeted policy interventions to facilitate broader adoption. Full article

Concrete arch dams represent a predominant dam type in water conservancy and hydropower projects in China. The control of concrete placement progress during construction directly impacts project quality and construction efficiency. Traditional manual monitoring methods, characterized by delayed response and strong subjectivity, struggle to meet the demands of modern intelligent construction management. This study introduces machine vision technology to monitor the concrete placement process and establishes an intelligent analysis system for construction scenes based on deep learning. By comparing the performance of U-Net and DeepLabV3+ semantic segmentation models in complex construction environments, the U-Net model, achieving an IoU of 89%, was selected to identify vibrated and non-vibrated concrete areas, thereby optimizing the concrete image segmentation algorithm. A comprehensive real-time analysis method for placement progress was developed, enabling automatic ternary classification and progress calculation for key construction stages, including concrete unloading, spreading, and vibration. In a continuous placement case study of Monolith No. 3 at a project site, the model’s segmentation results showed only an 8.2% error compared with manual annotations, confirming the method’s real-time capability and reliability. The research outcomes provide robust data support for intelligent construction management and hold significant practical value for enhancing the quality and efficiency of hydraulic engineering construction. Full article

The emergence of artificial intelligence as a transformative force in the field of information technology has exerted a significant impact on the development of renewable energy. In-depth analysis of the impact of AI on renewable energy development is crucial for promoting energy transition and facilitating sustainable development. This research utilizes a dataset comprising 30 provincial panels spanning from 2010 to 2023. This study found that AI technology can promote renewable energy development, a conclusion that still holds after robustness and endogeneity tests. An examination of the mechanism reveals that AI technology facilitates the advancement of renewable energy through the enhancement of trade openness and the concentration of manufacturing activities. The analysis of the moderating effect indicates that environmental regulation and environmental protection expenditures positively moderated the relationship between AI technology and renewable energy development and climate policy uncertainty negatively moderated the relationship between AI technology and renewable energy development. Further analysis revealed that AI technology has the potential to substantially improve the development of local renewable energy resources while also facilitating the advancement of renewable energy in adjacent areas, exhibiting spatial spillover effects. This study verifies the positive effects of AI technology on renewable energy development and enriches existing research perspectives in the field of energy economics. Full article

Farmers’ perceptions of improved rice varieties represent a critical initial step in their adoption as climate change adaptation strategies. This study examined farmers’ perceptions by integrating on-farm adaptive research, which compared the agronomic performance of rice varieties, with participatory approaches to capture farmers’ evaluation of improved varieties. A total of 81 farmers from climate-affected areas of Batang Regency were purposively selected as respondents. Data was collected through structured interviews and questionnaires administered during the evaluation of field demonstrations. Farmers’ perception levels were assessed using a Guttman scale and classified into three categories: high, medium, and low. Logistic regression analysis was subsequently employed to examine the relationship between farmers’ socio-demographic characteristics and their acceptance of improved rice varieties. The results indicate that, overall, farmers exhibited a low perception of improved rice varieties. Among the evaluated opinions, Inpari 32 HDB received the highest perception scores across all agronomic attributes. The regression results show that farm size and age significantly influence variety acceptance. The odds ratio for farm size (0.117) suggests that each additional hectare of cultivated land area reduces the likelihood of adopting improved rice varieties by approximately 88.3%, holding other factors constant. In contrast, the odds ratio for age (1.080) indicates that each additional year of age increases the probability of adoption by about 8%. Full article

In the context of industrial heritage conservation and adaptive reuse, the transformation of industrial buildings into patriotic education bases has emerged as a significant approach, where enhancing emotional education efficacy becomes crucial. This study adopts an evidence-based design (EBD) methodology, focusing on the Siwangzhang patriotic education base in Guangdong Province, to address the scientific evaluation and optimization of emotional arousal efficacy. The research rigorously follows the standardized EBD workflow: (1) during problem definition, the literature review establishes the dual objectives of quantitative assessment and spatial optimization; (2) evidence collection employs questionnaire surveys to capture emotional data from both static environmental nodes and dynamic activity nodes; (3) evidence analysis integrates descriptive analysis, factor analysis, emotional mapping visualization, and paired-sample t-tests. Key findings reveal the following: (1) spatial emotional distribution exhibits three distinct patterns—high-arousal clusters, single-node prominence areas, and emotional depressions; (2) dynamic training activities significantly enhance 66.7% of observed emotional variables. A seven-stage progressive training protocol was developed to achieve phased emotional cultivation. This study validates the applicability of EBD methodology in educational space optimization through a complete workflow, establishing an operational evaluation framework integrating spatial-behavioral-emotional metrics. It provides empirical evidence for targeted optimization of patriotic education bases while pioneering a data-driven transition from conventional experiential design. The results hold theoretical and practical significance for revitalizing industrial heritage through socially valuable functional transformations. Full article