Search Results (1,049)

The optimization capabilities of Kolmogorov–Arnold Networks (KANs) remain largely unexplored, which has limited their practical use in video anomaly recognition compared to conventional 3D-CNNs. To address this gap, we introduce a novel hybrid optimization framework that combines a Minimax Ant System (MMAS) for hyperparameter selection with a modified DARTS strategy for adaptive tuning of the 3D KAN architecture. Unlike existing approaches, our method simultaneously optimizes both learning dynamics and architectural configurations, enabling KANs to better exploit their expressive power in spatiotemporal feature extraction. Applied to a three-class video dataset, the proposed approach improved model accuracy to 87%, surpassing the performance of a standard 3D-CNN by 6%. Full article

To address the failure of unmanned aerial vehicle (UAV) target tracking caused by occlusion and limited field of view in dense low-altitude obstacle environments, this paper proposes a novel framework integrating occlusion-aware modeling and multi-UAV collaboration. A lightweight tracking model based on the Mamba backbone is developed, incorporating a Dilated Wavelet Receptive Field Enhancement Module (DWRFEM) to fuse multi-scale contextual features, significantly mitigating contour fragmentation and feature degradation under severe occlusion. A dual-branch feature optimization architecture is designed, combining the Distilled Tanh Activation with Context (DiTAC) activation function and Kolmogorov–Arnold Network (KAN) bottleneck layers to enhance discriminative feature representation. To overcome the limitations of single-UAV perception, a multi-UAV cooperative system is established. Ray intersection is employed to reduce localization uncertainty, while spherical sampling viewpoints are dynamically generated based on obstacle density. Safe trajectory planning is achieved using a Crested Porcupine Optimizer (CPO). Experiments on the Multi-Drone Multi-Target Tracking (MDMT) dataset demonstrate that the model achieves 84.1% average precision (AP) at 95 Frames Per Second (FPS), striking a favorable balance between speed and accuracy, making it suitable for edge deployment. Field tests with three collaborative UAVs show sustained target coverage in complex environments, outperforming traditional single-UAV approaches. This study provides a systematic solution for robust tracking in challenging low-altitude scenarios. Full article

Hydrologists need to predict extreme hydrological and meteorological events for design purposes, whose magnitude and probability are estimated using a probability distribution function (PDF). The choice of an appropriate PDF is crucial in describing the behavior of the phenomenon and the predictions can differ significantly depending on the PDF. So, the success of the probability distribution function in representing the data of extreme value series of natural events such as hydrology and climatology is of great importance. Depending on whether the series consists of maximum or minimum values, the theoretical probability density function must be appropriately fit to the right or left tail of the extreme data, which contains the most critical information. This study includes a combined evaluation of the performance of four different tests for selecting the appropriate probability distribution of maximum rainfall in Türkiye: Kolmogorov–Smirnov (KS) test, Anderson–Darling (AD) test, Probability Plot Correlation Coefficient (PPCC) test, and L-Moments Z^DIST test. Within the scope of the study, maximum rainfall series of seven rainfall durations from 15 to 1440 min, at rain gauge stations in 81 provinces of Türkiye, were examined. Goodness of fit was performed based on ranking using a combination of four different numerical tests (KS, AD, PPCC, Z^DIST). The probabilistic character of maximum rainfall was evaluated using a large dataset consisting of 567 time series with record lengths ranging from 45 to 80 years. The goodness of fit of distributions was examined from three different perspectives. The first is an examination considering rainfall durations, the second is a province-based examination, and the third is a general country-based assessment. In all three different perspectives, the Wakeby distribution was determined as the best fit candidate to represent the maximum rainfall in Türkiye. Full article

The aim of this study is to evaluate the mechanical properties and long-term stability of 3D-printable resins for permanent fixed dental prostheses (FDPs), focusing on whether material performance is influenced by 3D-printer type or by differences in resin formulations. Specimens (N = 621) were printed. CAD/CAM blocks (BRILLIANT Crios) served as control. Flexural strength (FS) with elastic modulus (E_calc), Weibull modulus (m), Martens’ hardness (HM), indentation modulus (EIT), elastic modulus (E_RFDA), shear modulus (G_RFDA), and Poisson’s Ratio (ν) were measured initially, after water storage (24 h, 37 °C), and after thermocycling (5–55 °C, 10,000×). SEM analysis assessed microstructure. Data were analyzed using Kolmogorov–Smirnov, ANOVA with Scheffe post hoc, Kruskal–Wallis with Mann–Whitney U, and Weibull statistics with maximum likelihood (α = 0.05). A ceramic crown printed with Midas showed higher FS, HM, and EIT values after thermocycling than with Pro55s, and higher E_calc scores across all aging regimes. A Varseo Smile Crown Plus printed with VarseoXS and AsigaMax showed a higher FS value than TrixPrint2, while AsigaMax achieved the highest initial E_calc and E_RFDA values, and VarseoXS did so after thermocycling. HM, EIT, and G_RFDA were higher for TrixPrint2 and AsigaMax printed specimens, while ν varied by system and aging. 3Delta Crown, printed with AsigaMax, showed the highest FS, E_calc, HM, EIT, and m values after aging. VarseoSmile triniQ and Bridgetec showed the highest E_RFDA and G_RFDA values depending on aging, and Varseo Smile Crown Plus exhibited higher ν initially and post-aging. Printer system and resin formulation significantly influence the mechanical and aging behaviors of 3D-printed FDP materials, underscoring the importance of informed material and printer selection to ensure long-term clinical success. Full article

The rapid expansion of AI applications in various domains demands models that balance predictive power with human interpretability, a requirement that has catalyzed the development of hybrid algorithms combining high accuracy with human-readable outputs. This study introduces a novel neuro-symbolic framework, OIKAN (Optimized Interpretable Kolmogorov–Arnold Network), designed to integrate the representational power of feedforward neural networks with the transparency of symbolic regression. The framework employs Gaussian noise-based data augmentation and a two-phase sparse symbolic regression pipeline using ElasticNet, producing analytical expressions suitable for both classification and regression problems. Evaluated on 60 classification and 58 regression datasets from the Penn Machine Learning Benchmarks (PMLB), OIKAN Classifier achieved a median accuracy of 0.886, with perfect performance on linearly separable datasets, while OIKAN Regressor reached a median R² score of 0.705, peaking at 0.992. In comparative experiments with ElasticNet, DecisionTree, and XGBoost baselines, OIKAN showed competitive accuracy while maintaining substantially higher interpretability, highlighting its distinct contribution to the field of explainable AI. OIKAN demonstrated computational efficiency, with fast training and low inference time and memory usage, highlighting its suitability for real-time and embedded applications. However, the results revealed that performance declined more noticeably on high-dimensional or noisy datasets, particularly those lacking compact symbolic structures, emphasizing the need for adaptive regularization, expanded function libraries, and refined augmentation strategies to enhance robustness and scalability. These results underscore OIKAN’s ability to deliver transparent, mathematically tractable models without sacrificing performance, paving the way for explainable AI in scientific discovery and industrial engineering. Full article

Objective: The aim of this study was to examine how transcranial electrical stimulation (tES) modulates intermuscular coherence (IMC) in sprinters and develop an interpretable neural network model for performance prediction. Methods: Thirty elite sprinters completed a randomized crossover trial involving three tES conditions: motor cortex stimulation (C1/C2), prefrontal stimulation (F3), and sham. Sprint performance metrics (0–100 m phase analysis) and lower-limb sEMG signals were collected. A Kolmogorov–Arnold Network (KAN) was trained to decode neuromuscular coordination–sprint performance relationships using IMC and time–frequency sEMG features. Results: Motor cortex tDCS increased 30–60 m sprint velocity by 2.2% versus sham (p < 0.05, η² = 0.25). γ-band IMC in key muscle pairs (rectus femoris–biceps femoris, tibialis anterior–gastrocnemius) significantly heightened under motor cortex stimulation (F > 4.2, p < 0.03). The KAN model achieved high predictive accuracy (R² = 0.83) through cross-validation, with derived symbolic equations mapping neuromuscular features to performance. Conclusions: Targeted tDCS enhances neuromuscular coordination and sprint velocity, while KAN provides a transparent framework for performance modeling in elite sports. Full article

We study a stochastic differential model for the dynamics of institutional corruption, extending a deterministic three-variable system—corruption perception, proportion of sanctioned acts, and policy laxity—by incorporating Gaussian perturbations into key parameters. We prove global existence and uniqueness of solutions in the physically relevant domain, and we analyze the linearization around the asymptotically stable equilibrium of the deterministic system. Explicit mean square bounds for the linearized process are derived in terms of the spectral properties of a symmetric matrix, providing insight into the temporal validity of the linear approximation. To investigate global behavior, we relate the first exit time from the domain of interest to backward Kolmogorov equations and numerically solve the associated elliptic and parabolic PDEs with FreeFEM, obtaining estimates of expectations and survival probabilities. An application to the case of Mexico highlights nontrivial effects: while the spectral structure governs local stability, institutional volatility can non-monotonically accelerate global exit, showing that highly reactive interventions without effective sanctions increase uncertainty. Policy implications and possible extensions are discussed. Full article

A modified Johnson–Mehl–Avrami–Kolmogorov (JMAK) model, including the heating rates, was proposed in this study to improve the accuracy of isothermal austenite formation kinetics prediction. Since the ultrafast heating process affects the behavior of ferrite recrystallization and austenite formation before the isothermal process, which in turn influences the subsequent isothermal austenite formation kinetics, the effects of varying austenitization temperatures and heating rates on isothermal austenite formation in cold-rolled quenching and partitioning (Q&P) steel, which remain insufficiently understood, were systematically investigated. Under a constant heating rate, the austenite formation rate initially increases and subsequently decreases as the austenitization temperature rises from formation start temperature A_c1 to finish temperature A_c3, and complete austenitization is achieved more quickly at elevated temperatures. At a given austenitization temperature, an increased heating rate was found to accelerate the isothermal transformation kinetics and significantly reduce the duration required to achieve complete austenitization. The experimental results revealed that both the transformation activation energy ($Q$) and material constant ($k_0$) decreased with increasing heating rates, while the Avrami exponent ($n$) showed a progressive increase, leading to the development of the heating-rate-dependent modified JMAK model. The model accurately characterizes the effect of varying heating rates on isothermal austenite formation kinetics, enabling kinetic curves prediction under multiple heating rates and austenitization temperatures and overcoming the limitation of single heating rate prediction in existing models, with significantly broadened applicability. Full article

Accurate prediction of CO₂ corrosion under dense-phase and supercritical conditions remains a critical challenge for oil and gas pipeline integrity management. While machine learning (ML) has been applied in this field, prevailing models like the Multilayer Perceptron (MLP) often struggle to capture the complex, non-linear interactions between multiple environmental parameters, limiting their predictive accuracy and robustness. To bridge this gap, this study innovatively introduces the Kolmogorov–Arnold Network (KAN) algorithm for CO₂ corrosion rate prediction. Utilizing a unique dataset of field-collected parameters (including dissolved O₂, H₂S, SO₂ concentrations, and water cut), we developed a KAN model and conducted systematic hyperparameter optimization. Our investigation revealed the optimal network configuration (3 layers, grid = 3) and, counterintuitively, that the steps parameter does not correlate positively with performance. Most significantly, comparative experiments demonstrated that the KAN model substantially outperforms traditional MLP, achieving superior prediction accuracy alongside faster computational speed and lower loss values. These findings not only provide a robust tool for precise corrosion prevention in oilfield operations but also highlight the potential of KAN as a novel, efficient, and highly accurate framework for tackling complex problems in materials degradation. Full article

While it is undisputed that Poisson GLMs represent the gold standard for counting COVID-19 deaths, recent studies have analyzed the seasonal growth and decline trends of these deaths in Italy using a simple segmented linear regression. They found that, despite an overall decreasing trend throughout the entire period analyzed (2021–2025), rising mortality trends from COVID-19 emerged in all summers and winters of the period, though they were more pronounced in winter. The technical reasons for the general unsuitability of using linear regression for the precise counting of deaths are well-known. Nevertheless, the question remains whether, under certain circumstances, the use of linear regression can provide a valid and useful tool in a specific context, for example, to highlight the slopes of seasonal growth/decline in deaths more quickly and clearly. Given this background, this paper presents a comparison between the use of linear regression and a Poisson GLM with the aforementioned death data, leading to the following conclusions. Appropriate statistical hypothesis testing procedures have demonstrated that the conditions of a normal distribution of residuals, their homoscedasticity, and the lack of autocorrelation were essentially guaranteed in this particular Italian case (weekly COVID-19 deaths in Italy, from 2021 to 2025) with very rare exceptions, thus ensuring the acceptable performance of linear regression. Furthermore, the development of a Poisson GLM definitively confirmed a strong agreement between the two models in identifying COVID-19 mortality trends. This was supported by a Kolmogorov–Smirnov test, which found no statistically significant difference between the slopes calculated by the two models. Both the Poisson and the linear model also demonstrated a comparably high accuracy in counting COVID-19 deaths, with MAE values of 62.76 and a comparable 88.60, respectively. Based on an average of approximately 6300 deaths per period, this translated to a percentage error of just 1.15% for the Poisson and only a slightly higher 1.48% for the linear model. Full article

The Kolmogorov–Arnold representation theorem states that any continuous multivariable function can be exactly represented as a finite superposition of continuous single-variable functions. Subsequent simplifications of this representation involve expressing these functions as parameterized sums of a smaller number of unique monotonic functions. Kolmogorov–Arnold Networks (KANs) have been recently proposed as an alternative to multilayer perceptrons. KANs feature learnable nonlinear activations applied directly to input values, modeled as weighted sums of basis spline functions. This approach replaces the linear transformations and sigmoidal post-activations used in traditional perceptrons. In this work, we propose a novel KAN variant by replacing both the inner and outer functions in the Kolmogorov–Arnold representation with weighted sinusoidal functions of learnable frequencies. We particularly fix the phases of the sinusoidal activations to linearly spaced constant values and provide a proof of their theoretical validity. We also conduct numerical experiments to evaluate its performance on a range of multivariable functions, comparing it with fixed-frequency Fourier transform methods, basis spline KANs (B-SplineKANs), and multilayer perceptrons (MLPs). We show that it outperforms the fixed-frequency Fourier transform B-SplineKAN and achieves comparable performance to MLP. Full article

With the rapid evolution of artificial intelligence technologies in power systems, data-driven time-series forecasting has become instrumental in enhancing the stability and reliability of power systems, allowing operators to anticipate demand fluctuations and optimize energy distribution. Despite the notable progress made by current methods, they are still hindered by two major limitations: most existing models are relatively small in architecture, failing to fully leverage the potential of large-scale models, and they are based on fixed nonlinear mapping functions that cannot adequately capture complex patterns, leading to information loss. To this end, an LLM-Empowered Kolmogorov–Arnold frequency learning (LKFL) is proposed for time series forecasting in power systems, which consists of LLM-based prompt representation learning, KAN-based frequency representation learning, and entropy-oriented cross-modal fusion. Specifically, LKFL first transforms multivariable time-series data into text prompts and leverages a pre-trained LLM to extract semantic-rich prompt representations. It then applies Fast Fourier Transform to convert the time-series data into the frequency domain and employs Kolmogorov–Arnold networks (KAN) to capture multi-scale periodic structures and complex frequency characteristics. Finally, LKFL integrates the prompt and frequency representations through an entropy-oriented cross-modal fusion strategy, which minimizes the semantic gap between different modalities and ensures full integration of complementary information. This comprehensive approach enables LKFL to achieve superior forecasting performance in power systems. Extensive evaluations on five benchmarks verify that LKFL sets a new standard for time-series forecasting in power systems compared with baseline methods. Full article

Neural networks have become a fundamental tool for solving complex problems, from image processing and speech recognition to time series prediction and large-scale data classification. However, traditional neural architectures suffer from interpretability problems due to their opaque representations and lack of explicit interaction between linear and nonlinear transformations. To address these limitations, Kolmogorov–Arnold Networks (KAN) have emerged as a mathematically grounded approach capable of efficiently representing complex nonlinear functions. Based on the principles established by Kolmogorov and Arnold, KAN offer an alternative to traditional architectures, mitigating issues such as overfitting and lack of interpretability. Despite their solid theoretical basis, practical implementations of KAN face challenges, such as optimal function selection and computational efficiency. This paper provides a systematic review that goes beyond previous surveys by consolidating the diverse structural variants of KAN (e.g., Wavelet-KAN, Rational-KAN, MonoKAN, Physics-KAN, Linear Spline KAN, and Orthogonal Polynomial KAN) into a unified framework. In addition, we emphasize their mathematical foundations, compare their advantages and limitations, and discuss their applicability across domains. From this review, three main conclusions can be drawn: (i) spline-based KAN remain the most widely used due to their stability and simplicity, (ii) rational and wavelet-based variants provide greater expressivity but introduce numerical challenges, and (iii) emerging approaches such as Physics-KAN and automatic basis selection open promising directions for scalability and interpretability. These insights provide a benchmark for future research and practical implementations of KAN. Full article

This work analyzes the behavior of the boundary layer subjected to stresses by obstacles using hourly measurements, in the form of time series, of meteorological variables (temperature (T), relative humidity (RH), and magnitude of the wind speed (WS)) in a given period. The study region is Santiago, the capital of Chile. The measurement location is in a rugged basin geography with a nearly pristine atmospheric environment. The time series are analyzed through chaos theory, demonstrating that they are chaotic through the calculation of the parameters Lyapunov exponent (λ > 0), correlation dimension (D_C < 5), Kolmogorov entropy (S_K > 0), Hurst exponent (0.5 < H < 1), and Lempel–Ziv complexity (LZ > 0). These series are simultaneous measurements of the variables of interest, before and after, of three different volumetric geometries arranged as obstacles: a parallelepiped, a cylinder, and a miniature mountain. The three geometries are subject to the influence of the wind and present the same cross-sectional area facing the measuring instruments oriented in the same way. The entropies calculated for each variable in each geometry are compared. It is demonstrated, in a first approximation, that volumetric geometry impacts the magnitude of the entropic fluxes associated with the measured variables, which can affect micrometeorology and, by extension, the climate in general. Furthermore, the study examines which geometry favors greater information loss or greater fractality in the measured variables. Full article

Background/Objectives: Orthodontic appliances disrupt oral biofilm homeostasis, leading to an increase in plaque and disease risk. Elastomeric modules (EMs) promote bacterial growth due to their material composition. Surface coatings have been developed to reduce bacterial colonization. We evaluated the mechanical, antimicrobial, and cell viability properties of a chitosan coating for EMs. Methods: EMs were coated with chitosan (CS) and chitosan-glutaraldehyde (CS-GTA) to assess antimicrobial and cell viability. Uncoated EMs were used as a control. These surface-coated modules were characterized and analyzed with Fourier transform infrared (FTIR) and Raman spectroscopy, and tensile testing. Antibacterial activity was assessed by colony-forming units (CFU) counts after incubation. Cell viability was tested with gingival fibroblasts using the MTT assay. ANOVA, Tukey, Kolmogorov–Smirnov, and Kruskal–Wallis tests were used for statistical analysis. Results: Raman spectra of the chitosan coatings showed characteristic molecular vibration bands. ANOVA revealed a significant difference in mechanical properties between the materials and between the control and the CS-GTA groups, confirmed by the Tukey post hoc test. No significant difference was observed between the groups in the Yield Stress test. All the coated groups showed reduced CFU counts in the antibacterial assay. The average cell viability of the coated groups was 85% and 89%. Conclusions: We synthesized CS and GTA-cross-linked chitosan coatings. The coatings did not affect the mechanical properties of the elastomeric modules. The chitosan and glutaraldehyde-cross-linked CS coatings inhibited bacterial growth. No significant differences were observed in antibacterial activity between the CS and the GTA-crosslinked chitosan coatings. Full article