Accurate Solutions Research Articles

A novel approach for ECG signal classification using sliding Euclidean quantization and bitwise pattern encoding.

This study aims to introduce a novel, computationally lightweight feature extraction technique called Sliding Euclidean Pattern Quantization (SEPQ), which encodes local morphological patterns of ECG signals using Euclidean distance-based binary representations within sliding windows. The proposed SEPQ method was evaluated using two ECG datasets. The first dataset contained three labeled classes (CHF, ARR, and NSR), while the second included four classes (ventricular beats (VB), supraventricular beats (SVB), fusion beats (FB), and NSR). Extracted features were classified using several machine learning models, with LightGBM achieving the highest performance-over 99% accuracy on the first dataset and above 93% on the second. A convolutional neural network (CNN) model was also employed for comparative analysis, both on raw data and in a hybrid configuration with SEPQ, yielding moderate yet noteworthy performance. Experimental results confirm that SEPQ offers a robust, interpretable, and highly accurate solution for ECG signal classification.

Integrated design of inductors using Monte Carlo tree search based on three-dimensional model

This paper proposes a scheme of the grouping Monte Carlo tree search for the optimization problems with broad branches. By classifying nodes which represent candidates of adjacent geometric parameters, the search goes through selecting and backpropagating based on the information of the groups, instead of nodes. Thus, the number of searching times can be reduced dramatically to traverse down leaf nodes, while ensuring the accuracy of the optimal solution. We have used it to successfully design a three-dimensional (3D) inductor.

Towards an AI tutor for undergraduate geotechnical engineering: a comparative study of evaluating the efficiency of large language model application programming interfaces

This study investigates the efficiency of large language model (LLM) application programming interfaces (APIs)—specifically GPT-4 and Llama-3—as AI tutors for undergraduate Geotechnical Engineering education. As educational needs in specialised fields like Geotechnical Engineering become increasingly complex, innovative teaching tools that provide personalised learning experiences are essential. Unlike previous studies on AI-driven education, our research uniquely focuses on assessing the role of retrieval-augmented generation (RAG) in improving the accuracy of LLM-generated solutions to Geotechnical problems. A dataset of 391 questions from the related textbook written by Das and Sobhan (Das B, Sobhan K. Principles of Geotechnical engineering, Eight Edition. In: Cengage Learning. 2014) was used for evaluation, with solutions sourced from the textbook’s manual. Performance benchmarking focused on 20 challenging questions previously identified by Chen et al. (Chen et al. in Geotechnics 4:470–498, 2024) as problematic for GPT-4 in Zero Shot tasks. GPT-4 with API support demonstrated superior accuracy, achieving accuracy rates of 95% at a temperature setting of 0.1, 82.5% at 0.5, and 60% at 1. In comparison, Llama-3 achieved an accuracy of 25% in Zero Shot tasks and 45% with API support at a temperature setting of 0.1. The findings highlight GPT-4’s potential as an AI tutor for Geotechnical Engineering education while demonstrating the need for domain-specific optimisation and advanced formula integration techniques. This study contributes to the ongoing discourse on AI in education by providing empirical evidence supporting the deployment of LLMs as personalised, adaptive teaching aids in engineering disciplines. Future work should explore optimised formula integration strategies, expanded domain knowledge bases, and long-term student learning outcomes.

Implementation of a Novel Gesture Recognition Technique for Real-Time Exercise Motion Detection

Human gesture and motion recognition systems have garnered major attention in recent years owing to their potential applications in fitness tracking, rehabilitation, and sports performance analysis. This research presents the implementation of a novel technique for the real-time detection and recognition of human gestures, specifically focusing on lower-body exercises such as squats. The proposed method leverages deep learning models combined with computer vision and motion capture technologies to accurately distinguish between correct and incorrect exercise forms. The system is trained using a dataset comprising annotated video recordings of various squat exercises, with key body landmarks extracted to track joint movements and detect posture anomalies. The core of the proposed technique involves a machine learning-based classification model that analyses the temporal and spatial features of human movement, providing corrective feedback to users. Gauge the model's performance utilising standard metrics like accuracy, and precision, and recall, along with F1-score, achieving an impressive accuracy rate of 97%, with high precision (95%), and recall (96%), and F1-score (95.5%). Moreover, a confusion matrix along with classification report are generated to gauge the model's effectiveness in distinguishing between correct and incorrect squat forms. This research adds to human motion detection by offering a robust, accurate, and scalable solution for real-time exercise correction, with potential applications in both fitness and rehabilitation domains.

CryoAlign2: efficient global and local Cryo-EM map retrieval based on parallel-accelerated local spatial structural features.

With the rapid advancements in Cryo-Electron Microscopy (Cryo-EM), an increasing number of high-resolution 3D density maps are being made publicly available, highlighting the urgent need for efficient structure similarity retrieval. Exploring map similarity at various levels is critical for fully utilizing these valuable resources. Our previously proposed CryoAlign can provide more accurate density map alignment while maintaining a low failure rate. However, CryoAlign only offers a method for aligning density maps, with low efficiency in local alignment, and has not yet been applied to the retrieval of Cryo-EM density maps. We have developed an alignment-based retrieval tool to perform both global and local retrieval. Our approach adopts parallel-accelerated CryoAlign for high-precision 3D alignment and transforms density maps into point clouds for efficient retrieval and storage. Additionally, a multi-dimension scoring function is introduced to accurately assess structural similarities between superimposed density maps. To demonstrate its applicability, we conducted thorough testing across different retrieval tasks, such as global, local or hybrid similarity retrieval. Our tool achieves up to a 7-fold speedup while supporting precise local alignments. Comprehensive experiments demonstrate that even when one density map is entirely contained within another, our tool performs exceptionally well in high-resolution density map retrieval. It provides researchers with an efficient and accurate solution for density map similarity search. The source code, documentation, and sample data can be downloaded at https://github.com/JokerL2/CryoAlign2. Supplementary data are available at Bioinformatics online.

Quasi-ray Tracing Method for Coulomb Interaction Estimation in Charged Particle Systems.

Coulomb interaction (CI) is a crucial factor contributing to the degradation of resolution in charged particle imaging systems. For example, certain inspection applications of scanning electron microscopes (SEM) require high probe current, which exacerbates resolution degradation due to CI. However, current methods for estimating CI, such as direct ray tracing and analytical approaches, struggle to balance speed and accuracy. This limitation significantly impedes the efficient optimization of SEM design parameters. In this study, we propose a quasi-ray tracing method based on perturbation theory, which simulates CI-induced trajectory displacement and the Boersch effect using a Monte Carlo ray tracing approach. Our method eliminates the need for iterative solvers employed in conventional direct ray tracing, thereby increasing simulation speed by over 105 times. Benchmarking against the direct ray tracing method demonstrates an error margin of less than 1% over a broad parameter range covering typical SEM applications. This quasi-ray tracing approach offers a fast, accurate, and general solution for CI estimation in charged particle imaging systems.

Application of the Tanh–Coth and HSI methods in deriving exact analytical solutions for the STFKS equation

The stochastic time-fractional Kuramoto–Sivashinsky (STFKS) equation models a wide range of physical phenomena involving spatio-temporal instabilities and noise-driven dynamics. Accurate analytical solutions to this equation are essential for understanding the influence of fractional derivatives and stochasticity in nonlinear systems. This study applies the Tanh–Coth method and He's Semi-Inverse (HSI) method in conjunction with the Truncated M-fractional derivative (TMFD) framework to derive exact solitary wave solutions of the STFKS equation. These approaches are implemented under a traveling wave transformation to reduce the governing partial differential equation to an ordinary differential equation. Exact analytical solutions are obtained for the STFKS equation, and graphical plots are presented to visualize the physical characteristics of the derived wave profiles under varying fractional orders and stochastic conditions. The results confirm that both the Tanh–Coth and HSI methods are effective and reliable for solving the STFKS equation. The graphical analysis reveals the significant impact of fractional parameters and stochastic terms on the solution behavior, demonstrating the practical utility of the proposed methodology in nonlinear stochastic modeling.

A Comparative Study of Solvers and Preconditioners for an SPE CO2 Storage Benchmark Reservoir Simulation Model

This study analyzes and evaluates the performance of various solvers and preconditioners for reservoir simulations of CO2 injection and long-term storage using the model 11B of SPE CSP (Society of Petroleum Engineers, 11th Comparative Solution Project) and the MATLAB Reservoir Simulation Toolbox (MRST). The SPE CSP 11 model serves as a benchmark for testing numerical methods for solving partial differential equations (PDEs) in reservoir simulations. The research focuses on the Biconjugate Gradient Stabilized (BiCGSTAB) and Loose Generalized Minimum Residual (LGMRES) solver methods, as well as multiple preconditioning techniques designed to improve convergence rates and reduce computational effort for CO2 storage. Extensive simulations were performed to compare the performance of different solver-preconditioner combinations under varying reservoir conditions, leveraging MRST’s flexible simulation capabilities. Key performance metrics, including iteration counts and computational time, were analyzed for the project. The results reveal trade-offs between computational efficiency and solution accuracy, providing valuable insights into the effectiveness of each approach. This study offers practical guidance for reservoir engineers and researchers seeking to analyze and optimize simulation workflows within MRST by identifying the strengths and limitations of specific solver-preconditioner combinations for complex linear systems.

Analytical study on steady seepage of a foundation pit adjacent to a structure

For a foundation pit adjacent to a structure with a common retaining wall, current mathematical models do not consider the interactions between seepage resulting from dewatering and the adjacent structures with the common retaining wall. In this study, the steady-state continuity equation of the seepage field considering an adjacent structure and retaining wall thickness is formulated by region dividing. The hydraulic head and water pressure are solved respectively using Fourier series expression and Bernoulli’s principle. A numerical model is established in FLAC3D to validate the accuracy of the analytical solution. The maximum deviation of analytical results is merely 2.76%, and the computation time is only two thousandths of that required by the numerical simulation. Further analysis is conducted on the effects of various engineering parameters on water pressure distribution, resultant water pressure and exit gradient, revealing that retaining wall thickness, adjacent structure width and upper/lower soil permeability ratio have non-negligible inhibitory effects on seepage, while pervious layer thickness exhibits a promotive effect and adjacent structure depth shows no influence on seepage. In conclusion, the proposed analytical solution can produce efficiently produce accurate results, making it valuable for practical applications in foundation engineering.

NeuroAdaptiveNet: A Reconfigurable FPGA-Based Neural Network System with Dynamic Model Selection

This paper presents NeuroAdaptiveNet, an FPGA-based neural network framework that dynamically self-adjusts its architectural configurations in real time to maximize performance across diverse datasets. The core innovation is a Dynamic Classifier Selection mechanism, which harnesses the k-Nearest Centroid algorithm to identify the most competent neural network model for each incoming data sample. By adaptively selecting the most suitable model configuration, NeuroAdaptiveNet achieves significantly improved classification accuracy and optimized resource usage compared to conventional, statically configured neural networks. Experimental results on four datasets demonstrate that NeuroAdaptiveNet can reduce FPGA resource utilization by as much as 52.85%, increase classification accuracy by 4.31%, and lower power consumption by up to 24.5%. These gains illustrate the clear advantage of real-time, per-input reconfiguration over static designs. These advantages are particularly crucial for edge computing and embedded applications, where computational constraints and energy efficiency are paramount. The ability of NeuroAdaptiveNet to tailor its neural network parameters and architecture on a per-input basis paves the way for more efficient and accurate AI solutions in resource-constrained environments.

An Improved Multi-Objective Grey Wolf Optimizer for Aerodynamic Optimization of Axial Cooling Fans

This paper introduces an improved multi-objective grey wolf optimizer (IMOGWO) and demonstrates its application to the aerodynamic optimization of an axial cooling fan. Building upon the traditional multi-objective grey wolf optimizer (MOGWO), several improvement strategies were adopted to enhance its performance. Firstly, the IMOGWO started population initialization based on the Bloch coordinates of qubits to ensure a high-quality initial population. Additionally, it employed a nonlinear convergence factor to facilitate global exploration and integrated the inspiration of Manta Ray Foraging to enhance the information exchange between populations. Finally, associative learning was leveraged for archive updating, allowing for perturbative mutation of solutions in crowded regions of the archive to increase solution diversity and improve the algorithm’s search capability. The proposed IMOGWO was applied to five multi-objective benchmark functions, comprising three two-objective and two three-objective problems, and experimental results were compared with three well-known multi-objective algorithms: the non-dominated sorting genetic algorithm II (NSGA II), MOGWO, and the multi-objective multi-verse optimizer (MOMVO). It is demonstrated that the proposed algorithm had advantages in convergence accuracy and diversity of solutions, which were quantified by the performance metrics (generational distance (GD), inverted generational distance (IGD), Spacing (SP), and Hypervolume (HV)). Furthermore, a multi-objective optimization process coupled with the IMOGWO algorithm and Computational Fluid Dynamics (CFD) was proposed. By optimizing the design parameters of an axial cooling fan, a set of non-dominated solutions was obtained within limited iteration steps. Consequently, the IMOGWO also presented an effective and practical approach for addressing multi-objective optimization challenges with respect to engineering problems.

Potential coefficient identification problem in parabolic equation with deep neural networks

Abstract The inverse problem of identifying an unknown space-dependent potential coefficient in the parabolic equation is considered from the additional observation at the terminal time in this work. A novel conditional stability estimate is established for a large terminal time T with suitable assumptions on the input data. Then the potential coefficient and solution of the parabolic equation are parameterized by separate deep neural networks (DNNs), and a new loss function is proposed to reconstruct the unknown potential coefficient. The DNN approximations of the potential coefficient for both continuous and empirical loss functions are analyzed rigorously via utilizing analogous arguments for the conditional stability. Meanwhile, the error estimates are expressed explicitly by the noise level and neural network architectural parameters, which yields a prior rule for determining the number of observations and choosing the size of neural networks. Some numerical experiments are provided to illustrate the robustness of the approach against various noise levels of measured observation and the accuracy of the numerical solutions.

Interpretable Deep Learning for Pediatric Pneumonia Diagnosis Through Multi-Phase Feature Learning and Activation Patterns

Pediatric pneumonia remains a critical global health challenge requiring accurate and interpretable diagnostic solutions. Although deep learning has shown potential for pneumonia recognition on chest X-ray images, gaps persist in understanding model interpretability and feature learning during training. We evaluated four convolutional neural network (CNN) architectures, i.e., InceptionV3, InceptionResNetV2, DenseNet201, and MobileNetV2, using three approaches—standard convolution, multi-scale convolution, and strided convolution—all incorporating the Mish activation function. Among the tested models, InceptionResNetV2, with strided convolutions, demonstrated the best performance, achieving an accuracy of 0.9718. InceptionV3 also performed well using the same approach, with an accuracy of 0.9684. For DenseNet201 and MobileNetV2, the multi-scale convolution approach was more effective, with accuracies of 0.9676 and 0.9437, respectively. Gradient-weighted class activation mapping (Grad-CAM) visualizations provided critical insights, e.g., multi-scale convolutions identified diffuse viral pneumonia patterns across wider lung regions, while strided convolutions precisely highlighted localized bacterial consolidations, aligning with radiologists’ diagnostic priorities. These findings establish the following architectural guidelines: strided convolutions are suited to deep hierarchical CNNs, while multi-scale approaches optimize compact models. This research significantly advances the development of interpretable, high-performance diagnostic systems for pediatric pneumonia using chest X-rays, bridging the gap between computational innovation and clinical application.

SmartCal: Calorie Estimation of Local Nepali Cuisine with Deep Learning-Powered Food Detection

Diet observation is a critical component of preventive healthcare, yet traditional manual calorie tracking methods are often inaccurate and time-consuming. This paper presents SmartCal, an automated system for calorie estimation of local Nepali cuisine using deep learning. The system focuses on foods like rice, lentils, spinach, and boiled eggs, which are staples in Nepali diets. It employs an ESP32 Camera Module to capture food images, which are processed by a Mask R-CNN (Region-Based Convolutional Neural Network) model trained specifically for Nepali cuisine. The model achieves a mean Average Precision at 50 (mAP50) of 99%, demonstrating high accuracy in food detection. Calorie estimation is performed based on standard cooking conditions and fixed portion sizes. Results are displayed via a web application, where users can set daily calorie targets and track their intake history. This system provides a practical, accurate, and user-friendly solution for dietary monitoring, particularly suited to the dietary habits of Nepali individuals.

DNS over HTTPS Tunneling Detection System Based on Selected Features via Ant Colony Optimization

DNS over HTTPS (DoH) is an advanced version of the traditional DNS protocol that prevents eavesdropping and man-in-the-middle attacks by encrypting queries and responses. However, it introduces new challenges such as encrypted traffic communication, masking malicious activity, tunneling attacks, and complicating intrusion detection system (IDS) packet inspection. In contrast, unencrypted packets in the traditional Non-DoH version remain vulnerable to eavesdropping, privacy breaches, and spoofing. To address these challenges, an optimized dual-path feature selection approach is designed to select the most efficient packet features for binary class (DoH-Normal, DoH-Malicious) and multiclass (Non-DoH, DoH-Normal, DoH-Malicious) classification. Ant Colony Optimization (ACO) is integrated with machine learning algorithms such as XGBoost, K-Nearest Neighbors (KNN), Random Forest (RF), and Convolutional Neural Networks (CNNs) using CIRA-CIC-DoHBrw-2020 as the benchmark dataset. Experimental results show that the proposed model selects the most effective features for both scenarios, achieving the highest detection and outperforming previous studies in IDS. The highest accuracy obtained for binary and multiclass classifications was 0.9999 and 0.9955, respectively. The optimized feature set contributed significantly to reducing computational costs and processing time across all utilized classifiers. The results provide a robust, fast, and accurate solution to challenges associated with encrypted DNS packets.

Reflecting topology consistency and abnormality via learnable attentions for airway labeling.

Accurate airway anatomical labeling is crucial for clinicians to identify and navigate complex bronchial structures during bronchoscopy. Automatic airway labeling is challenging due to significant anatomical variations. Previous methods are prone to generate inconsistent predictions, hindering preoperative planning and intraoperative navigation. This paper aims to enhance topological consistency and improve the detection of abnormal airway branches. We propose a transformer-based framework incorporating two modules: the soft subtree consistency (SSC) and the abnormal branch saliency (ABS). The SSC module constructs a soft subtree to capture clinically relevant topological relationships, allowing for flexible feature aggregation within and across subtrees. The ABS module facilitates interaction between node features and prototypes to distinguish abnormal branches, preventing the erroneous features aggregation between normal and abnormal nodes. Evaluated on a challenging dataset characterized by severe airway deformities, our method achieves superior performance compared to state-of-the-art approaches. Specifically, it attains an 83.7% subsegmental accuracy, along with a 3.1% increase in segmental subtree consistency, a 45.2% increase in abnormal branch recall. Notably, the method demonstrates robust performance in cases with airway deformities, ensuring consistent and accurate labeling. The enhanced topological consistency and robust identification of abnormal branches provided by our method offer an accurate and robust solution for airway labeling, with potential to improve the precision and safety of bronchoscopy procedures.

The Refined Physics-Informed Neural Networks for Nonlinear Convection-Reaction-Diffusion Equations Using Exponential Schemes

The nonlinear convection-reaction-diffusion equations model complex real-world phenomena across scientific and engineering disciplines. However, solving these equations analytically is often impossible due to their nonlinear nature. As a result, researchers have turned to numerical and computational methods to find approximate solutions. These methods, while effective, can struggle with issues such as stability, accuracy, and the ability to handle sharp gradients or complex interactions between convection, diffusion, and reaction terms. To address these challenges, this work introduces an enhanced Physics-Informed Neural Network (PINN) framework for convection-reaction-diffusion equations that incorporates exponential finite difference scheme residuals with the aim of enhancing solution accuracy and stability. To validate its performance, the framework has been tested on four well-known nonlinear partial differential equations: Burgers' Equation, Fisher's Equation, the Burgers-Huxley Equation, and the Newell-Whitehead-Segel Equation. The results obtained using the modified PINN framework are systematically compared with those obtained using traditional Physics-Informed Neural Networks and the Galerkin Finite Element Method. The comparisons reveal that the proposed framework consistently outperforms both approaches in terms of accuracy. These improvements highlight the effectiveness of integrating exponential finite difference scheme residuals into the PINN framework, making it a powerful and reliable tool for solving nonlinear convection-reaction-diffusion equations.

A Projection Method With Modular Grad‐Div Stabilization for Inductionless Magnetohydrodynamic Equations Based on Charge Conservation

ABSTRACTIn this paper, we proposed a fully discrete projection method with modular grad‐div stabilization for solving the time‐dependent inductionless magnetohydrodynamic equations. The method incorporates a minimally intrusive module into the classical projection method, serving as a postprocessing step, thereby enhancing solution accuracy and improving mass conservation. Additionally, a decoupled strategy is employed to separate the magnetic and fluid field functions from the original system. Since we only need to solve several linear subsystems at each time step, the numerical solutions can be obtained efficiently. In the spatial discretization, ‐conforming finite element pair is used to approximate the current density and electric potential, ensuring that the discrete current density is exactly divergence‐free. Therefore, the designed numerical scheme maintains the features of linearization, decoupling, unconditional energy stability, charge conservation, and improved mass conservation. The unconditional energy stability and convergence of the algorithm are rigorously analyzed and proven. Numerical results are presented to verify the robustness of the algorithm with respect to the stabilization parameters and to demonstrate the performance of the scheme, particularly in terms of its stability and accuracy.

Analytical Model and Gas Leak Source Localization Based on Acoustic Emission for Cylindrical Storage

A theoretical model is presented for the accurate detection of a gas leak source through a pinhole in a cylindrical storage vessel using the acoustic emission (AE) technique. Pinholes of various diameters ranging from 0.20 to 1.2 mm were installed as leak sources, and safe N2 was used as a filler gas. AE signals were measured and analyzed in terms of AE parameters (such as frequency, amplitude and RMS) as a function of angle and axial distance. Among them, the amplitude characteristic was the most important parameter to determine the leakage dynamics of AE with a continuous waveform. The simulation of AE amplitude was performed using the theoretical model for AE. For practical applications, the theoretical formula was modified into two semi-empirical equations by introducing the normalization method to fit the angular and axial characteristics of the observed AE amplitude, respectively. The main finding of this study is that the semi-empirical equations provide an accurate solution for leak source localization in the cylindrical vessel. As a priori knowledge, the value of in Green’s function, which determines the angular and axial dependence of the AE amplitude, was determined by applying external excitation to the cylinder surface. The proposed formulas provide a suitable approach for practical application in the localization of leak sources in cylindrical storage tanks.

Multiple strategy enhanced hybrid algorithm BAGWO combining beetle antennae search and grey wolf optimizer for global optimization

This study proposes BAGWO, a novel hybrid optimization algorithm that integrates the Beetle Antennae Search algorithm (BAS) and the Grey Wolf Optimizer (GWO) to leverage their complementary strengths while enhancing their original strategies. BAGWO introduces three key improvements: the charisma concept and its update strategy based on the sigmoid function, the local exploitation frequency update strategy driven by the cosine function, and the switching strategy for the antennae length decay rate. These improvements are rigorously validated through ablation experiments. Comprehensive evaluations on 24 benchmark functions from CEC 2005 and CEC 2017, along with eight real-world engineering problems, demonstrate that BAGWO achieves stable convergence and superior optimization performance. Extensive testing and quantitative statistical analyses confirm that BAGWO significantly outperforms competing algorithms in terms of solution accuracy and stability, highlighting its strong competitiveness and potential for practical applications in global optimization.