Search Results (830)

We introduce and study a class of multi-composite sigmoid neural network operators for Banach space-valued approximation. The proposed operators are generated by density-type kernels induced by finite compositions of seven standard sigmoid-type activation functions. The approximation is considered for continuous functions on compact intervals of the real line and on the whole real line, with values in an arbitrary Banach space $(X,\parallel ·\parallel )$. We prove quantitative pointwise and uniform convergence results by means of Jackson-type inequalities expressed through the first modulus of continuity. Higher-order and fractional approximation results are also obtained in terms of Banach space-valued derivatives and Caputo–Bochner fractional derivatives. The associated feed-forward neural network representation has one hidden layer and uses the multi-composite sigmoid function as its activation. Numerical experiments are presented to validate the theoretical estimates and to illustrate the approximation behavior of the proposed operators. In particular, we compare classical tanh-based operators, normalized self-composed activation operators, and heterogeneous multi-composite activation operators. The results show that self-composition and heterogeneous composition may improve the uniform approximation error for certain activation families and parameter choices, while also indicating that the observed improvement is activation-dependent and influenced by the composition order, kernel localization, and the regularity of the target function. Full article

Background/Objectives: Assessing inhibitory function and error detection is crucial for the early detection of age-related cognitive decline. This study aimed to investigate the neural network dynamics underlying these functions in younger and older adults to better understand age-related changes in cognitive control. Methods: We recorded electroencephalograms (EEGs) during an inhibitory control task in 17 older and 15 younger healthy adults. Behavioral performance was assessed, and directional functional connectivity was analyzed using Low-Resolution Electromagnetic Tomography (LORETA), isolated effective coherence (iCoh), and Full Vector Field analysis across the theta, alpha, and beta frequency bands. Results: Older adults showed significantly fewer correct responses than younger adults. During incorrect responses, older adults exhibited strong beta-band directionality from the ventral anterior cingulate cortex (ACC) to the left frontal polar cortex (FPC), alongside strong intra-ACC connectivity. During correct responses, they demonstrated alpha- and beta-band directionality from the left dorsolateral prefrontal cortex (DLPFC) to the right FPC. Conversely, compared with older adults, younger adults demonstrated significantly stronger mutual directionality within the ACC and widespread robust connectivity among the ACC, bilateral DLPFC, and FPC during correct responses. Conclusions: Efficient inhibitory control in older adults appears to rely on higher-order error-monitoring and error detection networks. The altered network dynamics in older adults suggest an age-related decline in immediate cognitive control. Evaluating these neural networks via EEGs provides a potential non-invasive biomarker for early cognitive decline and highlights higher-order executive control as a promising target for preventive interventions. Full article

Graph neural networks have attracted much attention and performed well in many downstream tasks. However, due to issues such as oversmoothing, existing graph neural networks are limited in their ability to quantitatively exploit higher-order neighborhood information. This paper introduces GAtD (Graph Attention Diffusion Method), which propagates attention to a wider range and aggregates higher-order information. We theoretically analyze the effectiveness of GAtD and demonstrate the convergence and linear complexity. A series of experiments demonstrates that, by combining diffusion and attention mechanisms, our method can effectively capture deep level relationships between nodes. Full article

The recent emergence of conversational Artificial Intelligence (AI) agents has profoundly transformed learning and teaching practices in higher education. These tools offer multiple advantages, ranging from cognitive assistance to enhanced student autonomy and efficiency. However, their actual impact on academic performance remains understudied, and the existing research often presents contradictory findings. To address this gap, the present study is the first to employ a Genetic Algorithm (GA) and Multi-Layer Perceptron Neural Networks (MLPNNs) to evaluate the influence of Generative AI Tools (GAITs) on students’ academic outcomes. A structured questionnaire was administered to 294 students from three Moroccan engineering schools in order to collect data on their use of these tools. An initial attempt to predict their grades using a statistical approach showed that familiarity with GAITs contributed positively to academic performance but had limited accuracy (39%), highlighting the need for more robust methods. Therefore, a hybrid model based on neural networks optimized with a GA was developed to better capture the complex relationships between the explanatory variables and academic performance. The results indicate that the GAIT-related variables considered in this study, taken in isolation, have a limited predictive capacity for students’ academic outcomes. This finding suggests that the available data does not capture the full complexity of the factors shaping academic success in contexts involving GAITs use. Full article

The widespread use of mobile Internet- and location-based services has generated large-scale check-in data in location-based social networks, creating opportunities for intelligent urban-mobility analysis and personalized mobility services. Making the next point-of-interest (POI) recommendation is an important task in this setting because it supports context-aware destination suggestion, travel assistance, and smart mobility services. However, existing methods still face challenges in jointly modeling higher-order mobility patterns, uneven time intervals, geographic reachability, and fine-grained intra-day temporal regularities. To address these issues, this paper proposes ASTHN, an Adaptive Spatio-Temporal Hypergraph Network for next POI recommendation. ASTHN constructs three fine-grained spatio-temporal context hypergraphs from minimum time interval, spatial proximity, and hourly preference, and uses hypergraph neural networks to learn view-specific POI representations. A context-adaptive fusion module then aligns and integrates multi-source spatio-temporal signals, while an ST-GRU with spatio-temporal gates captures dynamic trajectory evolution. Temperature scaling is further applied at the output layer to alleviate overly concentrated score distributions. Experiments on Foursquare-NYC and Foursquare-TKY show that ASTHN consistently outperforms representative baselines. With results reported as mean ± std over three random seeds, ASTHN improves over the strongest baseline by 3.79%, 14.62%, 2.28%, and 1.24% on NYC in Recall@5, Recall@10, NDCG@5, and NDCG@10, respectively. On TKY, the corresponding improvements are 5.83%, 37.20%, 13.86%, and 20.49%. Ablation, parameter, complexity, and application-oriented case analyses further demonstrate the effectiveness, stability, and practical usability of ASTHN for next POI recommendation in urban-mobility scenarios. Full article

Unmanned surface vessels (USVs) equipped with onboard vision are increasingly used in environmental monitoring, search and rescue, and autonomous navigation. However, conventional USV autonomy systems often adopt a decoupled design in which target perception and disturbance estimation are developed independently. Such systems may suffer performance degradation when visual observations become unreliable under water-surface reflections, illumination variations, or partial occlusions, while the disturbance observer still depends on manually tuned parameters under time-varying environmental disturbances. To address these issues, this paper proposes a three-stage co-optimized target perception and disturbance estimation framework for USVs. First, a lightweight hybrid convolutional neural network (CNN)–Transformer perception module is developed to extract robust vessel features under challenging water-surface visual conditions. Second, a reinforcement learning (RL)-driven mechanism is used to adaptively tune a higher-order sliding mode observer (HOSMO) for disturbance estimation. Third, a confidence-guided perception-observer co-optimization strategy is formulated, in which visual confidence is used to regulate observer adaptation and reduce estimation divergence during temporary perception degradation. Simulation and outdoor lake experiments demonstrate that the proposed framework improves visual matching accuracy, observer convergence, and estimation stability compared with conventional decoupled methods. The outdoor lake experiments provide initial real-world validation under natural illumination variations and mild water-surface disturbances, while further open-water and multi-vessel validation is planned for future work. Full article

Graph neural networks (GNNs) are increasingly used for medical imaging analysis and biological data modeling, where the integration of radiomics, topology, geometry, and generative artificial intelligence (AI) may improve representation learning from medical images and related biomedical data. Across the reviewed literature, GNNs show particular value for modeling spatial relationships, multimodal interactions, graph-structured biological networks, and non-Euclidean imaging features that are difficult to capture using conventional convolutional architectures alone. Topology- and geometry-aware approaches further expand this capability by encoding multi-scale structure, higher-order relationships, curvature, geodesic organization, and equivariant spatial priors. Hybrid graph–transformer models and generative graph methods represent emerging directions for modeling long-range dependencies, augmenting scarce datasets, supporting synthetic pretraining, and improving representation learning in low-label or heterogeneous biomedical settings. However, clinical translation remains limited by variability in graph construction, limited external validation, computational cost, scalability constraints, interpretability challenges, and uncertainty regarding the biological realism of synthetic data. Overall, this review highlights that GNN-based medical imaging analysis is most likely to advance when graph construction is biologically justified, model performance is evaluated across diverse clinical cohorts, and technical gains are paired with transparent validation, interpretability, and implementation strategies. Full article

Vertical axis wind turbines (VAWTs) have regained attention for distributed, urban, and floating offshore applications, yet the literature remains fragmented across competing rotor concepts and modelling traditions. This review consolidates the principal archetypes—Savonius, H-Darrieus, troposkein Darrieus, helical Darrieus, and Savonius–Darrieus hybrids—through five governing parameters: drag-versus-lift-driven operating principle, tip speed ratio $\mathsf{\lambda }=\mathsf{\omega }\mathrm{R}/{\mathrm{V}}_{\infty }$ (0.6–1.2 for Savonius; 2.5–5.0 for Darrieus), solidity $\mathsf{\sigma }=\mathrm{N}\mathrm{c}/\mathrm{R}$ (0.1–0.4), chord-based Reynolds number ${\mathrm{R}}_{\mathrm{e}\_\mathrm{c}}$ (${10}^5–{10}^6$), and peak power coefficient ${\mathrm{C}}_{\mathrm{p}\_\mathrm{m}\mathrm{a}\mathrm{x}}$ (0.15–0.25 for Savonius; 0.35–0.45 for optimized H-Darrieus). Off-design performance is dominated by unsteady mechanisms that quasi-steady streamtube models cannot resolve—leading edge vortex shedding, dynamic stall hysteresis, blade–wake interaction, and flow-curvature-induced virtual camber—each examined for its contribution to the instantaneous torque ${\mathrm{C}}_{\mathrm{T}}\left(\mathsf{\theta }\right)$ and the cycle-averaged ${\mathrm{C}}_{\mathrm{p}}$. Turbulence closures are benchmarked against phase-locked PIV and torque measurements: $\mathrm{k}–\mathsf{\omega }\mathrm{S}\mathrm{S}\mathrm{T}$ URANS captures peak-region ${\mathrm{C}}_{\mathrm{p}}$ to within $\pm 5–10\%$ but over-predicts torque below ${\mathsf{\lambda }}_{\mathrm{o}\mathrm{p}\mathrm{t}}$; the $\mathsf{\gamma }–{\mathrm{R}}_{\mathrm{e}\_\mathsf{\theta }}$ transition SST model reduces this error to $\pm 3–5\%$; DES, DDES, and LES reach $\pm 2–3\%$ at one to two orders of magnitude higher cost. Best practice computational fluid dynamics (CFD) guidelines are consolidated: domain extents of 15D upstream, 10D downstream, and 20D lateral; rotating sub-domain ${\mathrm{D}}_{\mathrm{r}\mathrm{o}\mathrm{t}}$ $\approx 1.5\mathrm{D}$; ${\mathrm{y}}^{+}\le 1$; $\Delta \mathsf{\theta }\le 0.1°$; and 20–30 revolutions before sampling. Performance enhancement strategies (variable pitch, guide vanes, helical twist, and hybridization) are reviewed quantitatively, with reported ${\mathrm{C}}_{\mathrm{p}}$ gains of $5–30\%$. Four research priorities are identified: (i) transition-sensitive turbulence closures validated below ${\mathrm{R}}_{\mathrm{e}\_\mathrm{c}}$ = $5\times {10}^5$; (ii) coupled aero-hydro-servo-elastic models for floating offshore VAWTs; (iii) machine-learning-augmented turbulence modelling—including physics-informed neural networks (PINNs) and neural-network-corrected RANS closures—to improve unsteady flow prediction at sub-LES cost; and (iv) integrated aeroacoustic–aeroelastic frameworks for urban and building-integrated deployment. Full article

The training of physics-informed neural networks (PINNs) for nonlinear multiphase flow in porous media is hampered by gradient conflicts between the individual components of the composite loss function. To address this problem, we propose a weighted gradient consistency metric that jointly accounts for the magnitudes and directions of the gradients of each loss term. Theoretical estimates of the convergence rate are derived, relating the proposed metric to the spectral properties of the preconditioner. The method is evaluated through a comparative study of optimizers—Adam, L-BFGS, and self-scaled Broyden—applied to three formulations of increasing complexity: a linear Buckley–Leverett model, a compressible two-phase model, and a fully nonlinear model with non-Newtonian rheology. The experiments demonstrate that self-scaled methods consistently achieve higher gradient alignment, faster loss reduction, and improved approximation accuracy compared to standard quasi-Newton and first-order baselines. Full article

Phishing and malicious websites are a serious threat on the internet. Attackers use fake links to trick users and steal their private information. Detecting these links is difficult because attackers change their tricks often. Many old methods cannot detect new or hidden threats. Some recent models use deep learning (DL), but they are large, slow, and hard to use in real-time systems. In this paper, we present a lightweight and accurate model called MiniLM-CNNLSTM. It combines a small transformer model (MiniLM) with a hybrid DL network using Convolutional Neural Network (CNN) and Long Short-Term Memory (LSTM) layers. The transformer learns the meaning of URLs. The CNN finds important patterns. The LSTM captures the order of characters. We also add handcrafted features that help the model detect tricky URLs. We test our method on two public datasets: the Phishing Site URLs dataset and the Malicious URLs dataset from Kaggle. We use 3-fold cross-validation and early stopping to ensure fair and stable results. The MiniLM-CNN-LSTM model outperformed previous benchmarks by achieving an average three-fold cross-validation accuracy of 98.98%, a precision of 98.63%, a recall of 98.29%, an F1-score of 98.46%, and a false positive rate of 0.68%. The proposed model has a higher accuracy, precision, recall, F1-score and a lower false positive rate, which enhances the accuracy by 1.88, precision by 3.77, recall by 4.17 and decreases the false positive rate by 61.58% compared with the strongest baseline (Distil BERT + CNN-LSTM), showing significant practical improvements. The results show that our approach is fast, small, and highly effective. It can detect phishing and malicious links with high accuracy. This makes it a good choice for real-time security systems like browsers, email filters, or firewalls. Full article

We propose the higher-order quasi-Newton (HOQN) method, a hybrid algorithm for unconstrained optimization that combines Newtonian predictors with higher-order correctors derived from vector extensions of the Traub, Chun, and Ostrowski methods, along with quasi-Newton updates of the inverse Hessian using Broyden–Fletcher–Goldfarb–Shanno (BFGS) or Davidon–Fletcher–Powell (DFP) formulas. We demonstrate that the resulting scheme achieves cubic local convergence order, representing a substantial improvement over the superlinear convergence typical of classical quasi-Newton methods, while maintaining a cost of $O\left(n^2\right)$ per iteration. We also analyze variants that incorporate two successive quasi-Newton updates, and show that they retain the same cubic order. Numerical experiments with the benchmark functions of Himmelblau and Freudenstein–Roth confirm the theoretical convergence order and show that the hybrid variants consistently require fewer iterations than BFGS, DFP, and Symmetric Rank-One (SR1). In the case of the Booth function, given its strictly convex quadratic structure, the proposed hybrid methods reach the global minimum in just two iterations and exhibit numerical accuracy superior to that of classical quasi-Newton methods. In addition, limited-memory variants (L-HOQN) are introduced; these are evaluated during the training of a convolutional neural network on the MNIST dataset, where they achieve test accuracies exceeding $99\%$ and outperform L-BFGS and standard stochastic gradient descent (SGD) at all tested learning rates. Full article

The integration of quantitative mathematical methods and artificial intelligence into agricultural monitoring systems represents a critical pathway toward data-driven decision-making in the contemporary precision agriculture economy. This study applies mathematical modeling and quantitative analysis to temporal NDVI (Normalized Difference Vegetation Index) raster datasets from six agricultural parcels in the Dobrogea region of Romania (2017 growing season), with the objective of supporting agronomic performance evaluation and operational decision-making. Higher-order statistical descriptors—variance, kurtosis, and skewness—were extracted from XML raster files and subjected to comprehensive visual analytics using kernel density estimation, three-dimensional surface modeling, and polynomial regression in Python. A feedforward Artificial Neural Network (ANN) with a 4-15-9-3-1 architecture was trained under four activation function and solver combinations (tanh/ReLU × Adam/SGD) to classify satellite sensing-date authenticity (is_sensing_date), a key data-quality indicator for operational crop monitoring workflows. Permutation-based feature importance analysis confirmed that variance is the dominant mathematical predictor (~35.8%), followed by kurtosis (~31.5%) and skewness (~26.6%), while the temporal month variable contributed least (~6.1%). The tanh–SGD configuration yielded the best training–test error balance for most individual datasets, while tanh–Adam performed optimally on the combined dataset. The inverse mathematical relationship between variance and kurtosis, and the direct co-variation between kurtosis and skewness, were consistent across all parcels, demonstrating the universality of these quantitative patterns in agricultural remote sensing data. These findings establish a replicable mathematical modeling framework applicable to predictive analytics, risk assessment of data quality, and performance evaluation in agricultural decision-making systems, with direct relevance to digital transformation strategies in the agri-economy sector. Full article

Background: Reaction time (RT) is a fundamental measure of information processing speed in cognitive neuroscience and is influenced by both structural and functional brain properties. While prior studies have independently linked white matter microstructure and EEG alpha oscillations to cognitive performance, their joint contribution to distinct aspects of RT remains unclear. This study aims to investigate whether multimodal data can dissociate neural systems underlying cognitive and motor components of processing speed. Methods: We analyzed diffusion tensor imaging, resting-state individual EEG alpha peak frequency (IAF), demographic variables, and behavioral RT measures from a GO/NO-GO paradigm in 24 healthy adults from the Cuban Human Brain Mapping Project. Behavioral metrics included the mean, standard deviation and skewness of reaction times for simple and complex tasks. Sparse multiple canonical correlation analysis was applied to identify multivariate associations across modalities. Results: Two significant latent dimensions were identified. The first dimension linked bilateral fronto-temporal association tracts (SLF, IFOF, UNC) with complex RT performance, reflecting higher-order cognitive processing. The second dimension associated motor and interhemispheric tracts (CGC, CST, ILF, forceps major and minor) with intra-individual asymmetric variability (skewness) across tasks, indicating a motor-execution consistency system. IAF did not significantly contribute to either dimension. Sex showed strong associations with both components. Conclusions: Distinct white matter networks were associated with separable cognitive and motor aspects of processing speed, while resting-state alpha frequency did not show stable contributions with behavioral variability in this sample. IAF showed minimal contribution within the identified sparse multivariate dimensions. These findings highlight the importance of multimodal and multivariate approaches for understanding and potentially disentangling complex brain–behavior relationships. Full article

This study presents some novel sharp estimates of the Euler–Maclaurin inequality using a new higher-order derivative Maclaurin identity. By utilizing the properties of convexity and classical inequalities, we exploit various novel tight boundaries of the Euler–Maclaurin inequality. They offer alternatives to measuring the sharp bounds of the mean integral of the higher-order differentiable mappings. In order to prove the importance and precision of the key findings, we apply graphical and numerical techniques. Another important section evaluates the behavior and validity of inequalities using a neural network model. The method is not only utilized to authenticate the results but also brings out the practical advancements of the study within a computational framework. The method and results of the article provide an insight and develop a solid connection between inequalities, higher-order derivative convex mappings, numerical analysis, approximation theory, and artificial neural networking. Full article

Background: The accuracy of tumor marker testing is critical for clinical decision-making. Patient-based real-time quality control (PBRTQC), as a complementary approach to traditional internal quality control (IQC), has been widely adopted in clinical laboratories. With the rapid advancement of automation and artificial intelligence (AI) in recent years, a large number of AI-based PBRTQC optimization algorithms have emerged. This study compared Patient-based real-time quality control integrating neural networks and joint probability analysis (NN-PBRTQC), Patient-Based Pre-Classified Real-Time Quality Control (PCRTQC), and traditional PBRTQC to identify the optimal method for quality control of tumor marker testing. Methods: The study utilized clinical tumor marker testing data from Peking University First Hospital. Six common tumor markers were selected, and constant error (CE) and proportional error (PE) were introduced as measures of analytical error. The False Alarm Rate (FAR) was used to reflect the specificity of the algorithms, while the Trimmed Average Number of Patient Results Affected Before Error Detection (tANPed) was used to reflect their sensitivity, in order to compare the clinical performance of the different models. Results: Under the same desired FAR (DFAR) of 0.1%, NN-PBRTQC reduced the tANPed for the six tumor markers by an average of 62% compared to the traditional PBRTQC while maintaining the same FAR, which demonstrated superior sensitivity of error detection. Meanwhile, although PCRTQC strictly controlled the FAR, its tANPed was 23% higher on average than that of the traditional PBRTQC, which indicated insufficient sensitivity of error detection. Conclusions: NN-PBRTQC demonstrated superior comprehensive quality control performance in the comparison of six common tumor markers. While ensuring that the FAR does not deviate from the DFAR, it significantly reduces tANPed, such that it could meet the specificity and sensitivity requirements of clinical testing. It is expected to enable more efficient and accurate detection of tumor marker errors. Full article