Search Results (3,060)

In the face of ever-evolving cyber threats, modern intrusion detection systems (IDS) must achieve long-term adaptability without sacrificing performance on previously encountered attacks. Traditional IDS approaches often rely on static training assumptions, making them prone to forgetting old patterns, underperforming in label-scarce conditions, and struggling with imbalanced class distributions as new attacks emerge. To overcome these limitations, we present a continual learning framework tailored for adaptive intrusion detection. Unlike prior methods, our approach is designed to operate under real-world network conditions characterized by high-dimensional, sparse traffic data and task-agnostic learning sequences. The framework combines three core components: a clustering-based memory strategy that selectively retains informative historical samples using DP-Means; multi-level knowledge distillation that aligns current and previous model states at output and intermediate feature levels; and a meta-learning-driven class reweighting mechanism that dynamically adjusts to shifting attack distributions. Empirical evaluations on benchmark intrusion detection datasets demonstrate the framework’s ability to maintain high detection accuracy while effectively mitigating forgetting. Notably, it delivers reliable performance in continually changing environments where the availability of labeled data is limited, making it well-suited for real-world cybersecurity systems. Full article

This study uses the CREATE^TM-AV/Kestrel simulation software to model a towed jumper scenario using standard aircraft settings to quantify paratrooper stability and risk of contact during static line airborne operations. The focus areas of this study include a review of the technical build-up, which includes aircraft, paratrooper and static line modeling, plus preliminary functional checkouts executed to verify simulation performance. This research and simulation development effort is driven by the need to meet the analysis demands required to support the US Army Personnel Airdrop with static line length studies and the North Atlantic Treaty Organization (NATO) Joint Airdrop Capability Syndicate (JACS) with airdrop interoperability assessments. Each project requires the use of various aircraft types, static line lengths and exit procedures. To help meet this need and establish a baseline proof of concept (POC) simulation, simulation setups were developed for a towed jumper from both the C-130J and C-17 using a 20-ft static line to support US Army Personnel Airdrop efforts. Concurrently, the JACS is requesting analysis to support interoperability testing to help qualify the T-11 parachute from an Airbus A400M Atlas aircraft, operated by NATO nations. Due to the lack of an available A400M geometry, the C-17 was used to demonstrate the POC, and plans to substitute the geometry are in order when it becomes available. The results of a nominal Computational Fluid Dynamics (CFD) simulation run using a C-17 and C-130J will be reviewed with a sample of the output to help characterize performance differences for the aircraft settings selected. The US Army Combat Capabilities Development Command Soldier Center (DEVCOM-SC) Aerial Delivery Division (ADD) has partnered with the US Air Force Academy (USAFA) High Performance Computing Research Center (HPCRC) to enable Modeling and Simulation (M&S) capabilities that support the Warfighter and NATO airdrop interoperability efforts. Full article

This study employs an improved fractional-order grey multivariable convolution model (FGMC(1,N,2r)) to predict abalone aquaculture output in Fujian, Shandong, and Guangdong. By integrating fractional-order accumulation (r₁, r₂) with a particle-swarm-optimization (PSO) algorithm, the model addresses limitations of handling multivariable interactions and sequence heterogeneity within small-sample regional datasets. Grey relational analysis (GRA) first identified key factors exhibiting the strongest associations with production: abalone production in Fujian and Shandong is predominantly influenced by funding for aquatic-technology extension (GRA degrees of 0.9156 and 0.8357, respectively), while in Guangdong, production was most strongly associated with import volume (GRA degree of 0.9312). Validation confirms that FGMC(1,N,2r) achieves superior predictive accuracy, with mean absolute percentage errors (MAPE) of 0.51% in Fujian, 3.51% in Shandong, and 2.12% in Guangdong, significantly outperforming benchmark models. Prediction of abalone production for 2024–2028 project sustained growth across Fujian, Shandong, and Guangdong. However, risks associated with typhoon disasters (X₆ and import dependency (X₅) require attention. The study demonstrates that the FGMC(1,N,2r) model achieves high predictive accuracy for regional aquaculture output. It identifies the primary drivers of abalone production: technology-extension funding in Fujian and Shandong, and import volume in Guangdong. These findings support the formulation of region-specific strategies, such as enhancing technological investment in Fujian and Shandong, and strengthening seed supply chains while reducing import dependency in Guangdong. Furthermore, by identifying vulnerabilities such as typhoon disasters and import reliance, the study underscores the need for resilient infrastructure and diversified seed sources, thereby providing a robust scientific basis for production optimization and policy guidance towards sustainable and environmentally sound aquaculture development. Full article

Generating non-standard fonts, such as running script (e.g., XingShu), poses significant challenges due to their high stroke continuity, structural flexibility, and stylistic diversity, which traditional component-based prior knowledge methods struggle to model effectively. While diffusion models excel at capturing continuous feature spaces and stroke variations through iterative denoising, they face critical limitations: (1) style leakage, where large stylistic differences lead to inconsistent outputs due to noise interference; (2) structural distortion, caused by the absence of explicit structural guidance, resulting in broken strokes or deformed glyphs; and (3) style confusion, where similar font styles are inadequately distinguished, producing ambiguous results. To address these issues, we propose a novel skeleton-guided diffusion model with three key innovations: (1) a skeleton-constrained style rendering module that enforces semantic alignment and balanced energy constraints to amplify critical skeletal features, mitigating style leakage and ensuring stylistic consistency; (2) a cross-scale skeleton preservation module that integrates multi-scale glyph skeleton information through cross-dimensional interactions, effectively modeling macro-level layouts and micro-level stroke details to prevent structural distortions; (3) a contrastive style refinement module that leverages skeleton decomposition and recombination strategies, coupled with contrastive learning on positive and negative samples, to establish robust style representations and disambiguate similar styles. Extensive experiments on diverse font datasets demonstrate that our approach significantly improves the generation quality, achieving superior style fidelity, structural integrity, and style differentiation compared to state-of-the-art diffusion-based font generation methods. Full article

Steel products play an important role in daily production and life as a common production material. Currently, the quality of steel products is judged by manual experience. However, various inspection criteria employed by human operators and complex factors and mechanisms in the steelmaking process may lead to inaccuracies. To address these issues, we propose a learning-based method for steel quality prediction, which is named DRBoost,based on multiple machine learning techniques, including Decision tree, Random forest, and the LSBoost algorithm. In our method, the decision tree clearly captures the nonlinear relationships between features and serves as a solid baseline for making preliminary predictions. Random forest enhances the model’s robustness and avoids overfitting by aggregating multiple decision trees. LSBoost uses gradient descent training to assign contribution coefficients to different kinds of raw materials to obtain more accurate predictions. Five key chemical elements, including carbon, silicon, manganese, phosphorus, and sulfur, which significantly influence the major performance characteristics of steel products, are selected. Steel quality prediction is conducted by predicting the contents of these chemical elements. Multiple models are constructed to predict the contents of five key chemical elements in steel products. These models are symmetrically complementary, meeting the requirements of different production scenarios and forming a more accurate and universal method for predicting the steel product’s quality. In addition, the prediction method provides a symmetric quality control system for steel product production. Experimental evaluations are conducted based on a dataset of 2012 samples from a steel plant in Liaoning Province, China. The input variables include various raw material usages, while the outputs are the content of five key chemical elements that influence the quality of steel products. The experimental results show that the models demonstrate their advantages in different performance metrics and are applicable to practical steelmaking scenarios. Full article

In this paper, based on Generalized Autoregressive Conditional Heteroskedasticity (GARCH) model, five financial serials are dynamically weighted, and then China’s Financial Conditions Index is synthesized to measure China’s financial cycle. After that, using the monthly data of 2000–2023 as sample space, this paper utilizes the Markov Switching (MS) model to analyze the characteristics of China’s financial cycle and to investigate the four-zone system. Then, the Vector Autoregression (VAR) model focuses on investigating the macroeconomic effects of China’s financial cycle. The findings are as follows: Firstly, the dynamic weighting approach based on GARCH model is more suitable for valuating China’s financial cycle. Secondly, China’s financial cycle has a strong inertia at the state of transition and the imbalance of China’s overall financial situation is very common. Additionally, China’s financial cycle is distinctly characterized by the double asymmetry of fewer contractions and more expansions, shorter expansions, and longer expansions. Thirdly, China’s financial expansion offers a nine-month short-term stimulus to output and exerts lasting upward pressure on prices. Full article

The Yellow River (YR) has the highest suspended sediment concentration in the world, with its water and sediment exerting a significant influence on nutrient transport and transformation processes. The periodic regulation of water and sediment by the Xiaolangdi Dam, has significantly altered downstream water and sediment transport. This study examined the impact of the Xiaolangdi Dam’s 2023 water-sediment regulation on nitrogen dynamics in the lower Yellow River (LYR). Surface water, suspended sediment, and deposited sediment samples were collected at seven downstream stations to analyze changes in nitrogen concentration, sources, and transformation processes. As the water regulation stage progresses, the (total nitrogen) TN concentration in the water phase decreased, while that of NO₃^--N increased slightly. Concurrently, the inorganic nitrogen concentration in the suspended phase also declined. As the sediment regulation stage progresses, the TN and NO₃⁻-N concentrations in the water phase continued to decrease, while the inorganic nitrogen concentration in the suspended phase showed an initial increase followed by a decrease. As the early stage of sediment regulation progresses, ammonia concentrations decreased, while nitrate concentrations increased and δ¹⁸O-NO₃⁻ value decreased indicated nitrification occurred. As the late stage of sediment regulation progresses, nitrate concentrations decreased and the δ¹⁵N-NO₃⁻ value increased, indicated denitrification occurred. The TN flux during water-sediment regulation reaches 41.5 kt (14.6% of the annual flux). During the water-sediment regulation stage, the main nitrate sources were manure and sewage. This contribution peaked at 54.2% during the sediment regulation stage. The research results provide a scientific basis for the relationship between water and sediment changes and nitrogen output changes in the LYR. Full article

Organismal biosensing leverages the olfactory acuity of living systems to detect volatile organic compounds (VOCs) associated with cancer, offering a low-cost and non-invasive complement to conventional diagnostics. Early studies demonstrate its feasibility across diverse platforms. In C. elegans, chemotaxis assays on urine samples achieved sensitivities of 87–96% and specificities of 90–95% in case–control cohorts (n up to 242), while calcium imaging of AWC neurons distinguished breast cancer urine with ~97% accuracy in a small pilot cohort (n ≈ 40). Trained canines have identified prostate cancer from urine with sensitivities of ~71% and specificities of 70–76% (n ≈ 50), and AI-augmented canine breath platforms have reported accuracies of ~94–95% across ~1400 participants. Insects such as locusts and honeybees enable ultrafast neural decoding of VOCs, achieving 82–100% classification accuracy within 250 ms in pilot studies (n ≈ 20–30). Collectively, these platforms validate the principle that organismal behavior and neural activity encode cancer-related VOC signatures. However, limitations remain, including small cohorts, methodological heterogeneity, and reliance on binary outputs. This review proposes a Dual-Pathway Framework, where Pathway 1 leverages validated indices (e.g., the Chemotaxis Index) for high-throughput screening, and Pathway 2 applies machine learning to high-dimensional behavioral vectors for cancer subtyping, staging, and monitoring. By integrating these approaches, organismal biosensing could evolve from proof-of-concept assays into clinically scalable precision diagnostics. Full article

Tree rings, tree needles, and moss can be used as biomonitors to evaluate atmospheric pollutant concentrations and deposition patterns spanning different timescales. This study compares output from air quality modeling and measurements to patterns observed using a combination of sulfur concentration and isotope composition in moss (using moss bags and controls) as biomonitors in a region of southern Alberta, Canada influenced by industrial emissions. Tree rings allow comparisons of historical to current sulfur deposition patterns. Moss, which integrates atmospheric nutrients during growth, allows for concurrent comparisons. The contrast of inorganic and organic sulfur within conifer tree needles provides a measure of pollutant uptake over their short lifespans. Sulfur uptake within biomonitors in a southern Alberta ecosystem allow assessment of the presence (in moss, needles) and effects (on conifer growth) of atmospheric sulfur deposition from industrial emissions. These data were examined relative to California Puff (CALPuff) model projections and traditional active and passive air quality sampling. Patterns in sulfur isotope abundance (δ³⁴S) from moss bags placed throughout the eastern slopes of the southern Alberta foothills of the Rocky Mountains implicate local industry as the dominant atmospheric sulfur source over winter, with the tissues of conifers (needles and cores) and moss decreasing with distance from industrial emissions. This was consistent with apportionment calculations based on active and passive sampling, which also showed a surprising trend of sulfur deposition upwind of the industrial stack in the mountains to the west. δ³⁴S values for pine needles and tree rings were consistent with greater sulfur stress and reductions in tree growth associated with increased industrial sulfur concentrations and deposition. We conclude that plant biomonitors are effective short-term (tree needles and moss) and long-term (tree cores) indicators of sulfur pollution in a complex, mountainous landscape. Full article

Multi-wavelength random lasers with switchable modes have advantages in the fields of novel light source and information security. Here, we propose a dual-wavelength phase transition random laser, which can modulate lasing modes arbitrarily assisted by the phase transition hydrogel. Once the phase transition occurs in hydrogel, the scattering properties of light in the random system changes, affecting the optical feedback mechanism and enabling reversible switching of the dual-wavelength random laser mode between incoherent and coherent states. More appealing, random lasing mixed incoherent mode and coherent mode have been obtained for the first time by controlling the local phase transition of the sample. Based on these properties, an information encryption system is constructed by encoding spectral fingerprints at different modes. This work provides an effective way to precisely control the output modes at different wavelengths in the multi-wavelength random laser, further expanding the application of random lasers in multifunctional light sources, color imaging, and information safety. Full article

Malignant pleural mesothelioma (MPM) is a rare and aggressive malignancy associated with occupational or environmental exposure to asbestos. Effective management of MPM remains challenging due to its prolonged latency period and the typically late onset of clinical symptoms. Accordingly, there is an increasing demand for the implementation of reliable, non-invasive, and data-driven diagnostic strategies within large-scale screening programs. In this context, the chemical profiling of volatile organic compounds (VOCs) in exhaled breath has recently gained recognition as a promising and non-invasive approach for the early detection of cancer, including MPM. Therefore, in this cross-sectional observational study, an overall number of 125 individuals, including 64 MPM patients and 61 healthy controls (HC), were enrolled. End-tidal breath fraction (EXP) was collected directly onto two-bed adsorbent cartridges by an automated sampling system and analyzed by thermal desorption–gas chromatography–mass spectrometry (TD-GC/MS). A machine learning approach based on a random forest (RF) algorithm and trained using a 10-fold cross-validation framework was applied to experimental data, yielding remarkable results (AUC = 86%). Fifteen VOCs reflecting key metabolic alterations characteristic of MPM pathophysiology were found to be able to discriminate between MPM and HC. Moreover, twenty breath samples from asymptomatic former asbestos-exposed (AEx) and eight MPM patients during follow-up (FUMPM) were exploratively analyzed, processed, and tested as blinded samples by the validated statistical method. Good agreement was found between model output and clinical information obtained by CT. These findings underscore the potential of breath VOC analysis as a non-invasive diagnostic approach for MPM and support its feasibility for longitudinal patient and at-risk subjects monitoring. Full article

Background: Large language models (LLMs) such as Google Gemini have demonstrated strong capabilities in natural language generation, but their ability to create medical assessment items aligned with National Board of Medical Examiners (NBME) standards remains underexplored. Objective: This study evaluated the quality of Gemini-generated NBME-style pharmacology questions using a structured rubric to assess accuracy, clarity, and alignment with examination standards. Methods: Ten pharmacology questions were generated using a standardized prompt and assessed independently by two pharmacology experts. Each item was evaluated using a 16-criterion NBME rubric with binary scoring. Inter-rater reliability was calculated (Cohen’s Kappa = 0.81) following a calibration session. Results: On average, questions met 14.3 of 16 criteria. Strengths included logical structure, appropriate distractors, and clinically relevant framing. Limitations included occasional pseudo-vignettes, cueing issues, and one instance of factual inaccuracy (albuterol mechanism of action). The evaluation highlighted Gemini’s ability to produce high-quality NBME-style questions, while underscoring concerns regarding sample size, reproducibility, and factual reliability. Conclusions: Gemini shows promise as a tool for generating pharmacology assessment items, but its probabilistic outputs, factual inaccuracies, and limited scope necessitate caution. Larger-scale studies, inclusion of multiple medical disciplines, incorporation of student performance data, and use of broader expert panels are recommended to establish reliability and educational applicability. Full article

For accurate and reliable monitoring, compliance monitoring devices (CMDs) in Port State Control must meet strict and uniform quality standards. This study evaluates how effectively CMDs, using variable fluorescence (VF) and fluorescein diacetate (FDA) technologies, detect live organisms in the 10–50 μm size range. Employing a detailed analytical framework, we analyzed key performance indicators, including accuracy, precision, sensitivity, specificity, trueness, detection limits, and reliability by comparing CMD outputs to those of traditional microscopic methods. Reliability assessments revealed that VF-type CMD and FDA-type CMD performed robustly, with a stability rate of 99% for both, surpassing the 90% verification threshold. Precision analysis indicated an average CV exceeding 0.25; however, some samples, especially those below the D-2 standard, achieved a CV of less than 0.25. Concordance evaluations revealed that VF-CMDs and FDA-CMDs achieved rates of 63% and 55%, respectively, falling short of the 80% verification standard and underscoring the need for further calibration or optimization. Structural equation modeling shows that organism density significantly influences CMD performance. These findings underscore the challenges of accurately detecting low organism concentrations, further complicated by biological diversity and environmental variability. Despite their limitations in assessing ballast water compliance, CMDs are effective initial screening tools. Full article

A variety of factors, such as absorption, reflection, and attenuation by atmospheric elements, influence the quantity of solar energy that reaches the surface of the Earth. This, in turn, impacts photovoltaic (PV) power generation. In light of this, a digital assessment of solar energy variability through short-term measurements was conducted to enhance PV power output. The clear-sky index $\left(K_t^{*}\right)$ methodology was employed, effectively eliminating any indications of solar energy obstruction and comparing the measured radiation to the theoretical clear-sky radiation. The solar energy data were gathered in Mozambique, specifically in the southern region at Maputo–1, Massangena, Ndindiza, and Pembe, in the mid-region at Chipera, Nhamadzi, Barue–1, and Barue–2, as well as in the northern region at Nipepe-1, Nipepe-2, Nanhupo-1, Nanhupo-2, and Chomba, over the period from 2005 to 2024, with measurement intervals ranging from 1 to 10 min and 1 h during the measurement campaigns conducted by FUNAE and INAM, with additional data sourced from the PVGIS, Meteonorm, NOAA, and NASA solar databases. The analysis indicates a $K_t^{*}$ value with a density approaching 1 for clear days, while intermediate-sky days exhibit characteristics that lie between those of clear and cloudy days. It can be inferred that there exists a robust correlation among sky types, with values ranging from 0.95 to 0.89 per station, alongside correlated energies, which experience a regression with coefficients between 0.79 and 0.95. Based on the analysis of the sample, the region demonstrates significant potential for solar energy utilization, and similar sampling methodologies can be applied in other locations to optimize PV output and other solar energy projects. Full article

A Bayesian approach for constructing ARMA probability density estimators is proposed. Such estimators are ratios of trigonometric polynomials and have a number of advantages over Fourier series estimators, including parsimony and greater efficiency under common conditions. The Bayesian approach is carried out via MCMC, the output of which can be used to obtain probability intervals for unknown parameters and the underlying density. Finite sample efficiency and methods for choosing the estimator’s smoothing parameter are considered in a simulation study, and the ideas are illustrated with data on a wine attribute. Full article