Abstract Density Peak Clustering (DPC) is a widely used clustering method that automatically identifies cluster centers based on local density and distance metrics. However, DPC tends to converge to local optimal solutions when processing datasets with uneven density distributions. This paper proposes DPC-DIST, an improved algorithm that addresses this limitation through a geometric distribution coefficient δ . By incorporating spatial distribution characteristics into density calculations, DPC-DIST effectively prevents convergence to suboptimal solutions. Comprehensive experiments on six benchmark datasets (Wine, Iris, Seeds, Blobs, R15, R3) demonstrate that DPC-DIST consistently outperforms the original DPC, K-means++, and spectral clustering algorithms. Significant improvements were observed in high-dimensional data processing, with 9.3 % increase in Silhouette Coefficient and 47.7 % improvement in Calinski-Harabasz Score on the Wine dataset. The algorithm shows particular advantages in facility location optimization and logistics planning applications, demonstrating its practical value for real-world scenarios.

Abstract Within this study, we delve into novel optical solitons concerning the perturbed nonlinear Schrödinger equation, which governs the propagation of electromagnetic waves in magneto-optic waveguides. These components play vital roles in optical transmission lines and lasers, serving critical functions within integrated optical circuits as circulators, modulators, and isolators. Employing the simplest equation method, the 1 φ ( δ ) , φ ′ ( δ ) φ ( δ ) \left(\phantom{\rule[-0.75em]{}{0ex}},\frac{1}{\varphi \left(\delta )},\frac{\varphi ^{\prime} \left(\delta )}{\varphi \left(\delta )}\right) method, and the generalized Riccati equation mapping method, we derive single and multi-optical soliton solutions. Our analysis yields a diverse array of solutions, including multi-soliton, singular, combo bright-dark, periodic singular, bright, and dark optical solitons. Additionally, we discuss the parametric conditions crucial for both the existence and shaping of these solitons. The visual representation of our findings through three-dimensional, two-dimensional, and contour plots elucidates the dynamic phenomena and interprets the physical implications of these solutions using various parameter values. This research demonstrates the efficacy of rational analytical methods in elucidating the intricate dynamics of soliton solutions in nonlinear optical systems, thereby offering valuable insights for further exploration in the fields of physics, mathematics, sciences, and engineering.

Abstract T-node tensile specimens, including both rectangular steel pipes and rectangular steel pipe concrete configurations, were developed based on the results of tensile tests conducted on concrete-filled rectangular steel pipe T-nodes. The specimens incorporated Perfobond Leiste (PBL) stiffeners, with concrete-filled rectangular steel pipes used for the main pipes and square steel pipes as the branch (stub) pipes. The fatigue performance of the concrete-filled rectangular steel pipe T-nodes under tensile loading, in-plane bending, and out-of-plane bending was investigated using PBL stiffeners as internal reinforcement. To determine the stress concentration parameters of the branch pipes, a nonlinear finite element analysis was performed using ABAQUS (a finite element analysis software suite developed by Dassault Systèmes) software. The analysis adopted a width-to-thickness ratio of 0.4 for the branch pipe and 27 for the main pipe. The results indicate that the hot spot locations of the PBL-stiffened concrete-filled rectangular steel pipe T-nodes are consistent with those of both the unstiffened rectangular steel pipe nodes and the concrete-filled nodes. Furthermore, the stress concentration factor of the PBL-stiffened node is significantly lower than that of the unstiffened steel pipe node, indicating markedly improved fatigue resistance.

Abstract With the increasing emphasis on environmental protection in various regions, reducing electricity emissions for enterprises has become a popular development trend. The research aims to design a machine learning-based power data warning method to assist enterprises in reducing emissions and controlling costs. In terms of research methods, load type analysis is first conducted for industrial enterprises, followed by the introduction of long short-term memory (LSTM) networks to build a basic model for power data prediction. Support vector machines are used to optimize the model’s large sample requirements, and the two models are integrated to improve accuracy. Key research has found that the proposed model performed excellently. The minimum relative prediction error was 0.20%, the maximum error fluctuation was 0.78%, the accuracy was 12.85% higher than the LSTM model, and the recall was 11.60% higher. On 220 kV, the testing time was 17.5% faster than the data prediction model and 36.0% faster than the multi-task learning model, and the accuracy was always the best. Simulation experiments showed that after data warning, carbon emissions could be reduced by up to 48.26%, and electricity costs could be reduced by up to 60.48%. The machine learning-based power data warning method proposed in this study has important practical application value and can effectively help enterprises achieve emission reduction and cost control goals.

Abstract This study addresses the accuracy challenges in e-commerce sentiment classification and thus provides valuable insight for businesses to enrich strategies toward the interpretation of customer feedback and improvement of product development. This article elaborately contrasts long short-term memory (LSTM)-based models with traditional machine learning models, like support vector machines (SVM), random forest, and Naive Bayes classifiers. The authors have used a large dataset of customer reviews from famous e-commerce websites, pre-processed for noise reduction and standardization of the input. Our LSTM models were implemented using state-of-the-art deep learning frameworks, with special consideration while performing hyperparameter tuning. The results showed that the bidirectional LSTM model had the best performance of 92.1% in accuracy, 91.8% in precision, 92.0% in recall, and an F1-score of 91.9%. In comparison, traditional machine learning approaches gave accuracies: 86.5% for SVM, 85.1% for random forest, and 82% for Naive Bayes. Hyperparameter tuning returned the best configurations for the LSTM models: 128 LSTM units, a dropout of 0.3, with learning rates 0.001 for LSTM and 0.0001 for bidirectional LSTM. These optimizations contributed hugely to the performance improvements observed in these models. Error analysis gave insights into the challenges that the models faced. Sarcasm and irony accounted for 22% of the classification errors, while mixed sentiment accounted for 18%, and implicit accounted for 15%. To sum up, this research has shown the efficiency of LSTM models on e-commerce user review sentiment analysis, especially bidirectional LSTM. These models show a much better result in sentiment classification compared to traditional machine learning techniques, which proves them to have huge potential for improving the accuracy of the task in real-world applications.

Abstract Sports videos contain a large number of irrelevant backgrounds and static frames, which affect the efficiency and accuracy of temporal action detection. To optimize sports video data processing and temporal action detection, an improved multi-level spatiotemporal transformer network model is proposed. The model first optimizes the initial feature extraction of videos through an unsupervised video data preprocessing model based on deep residual networks. Subsequently, multi-scale features are generated through feature pyramid networks. The global spatiotemporal dependencies of actions are captured by a spatiotemporal encoder. The frame-level self-attention module further extracts keyframes and highlights temporal features, thereby improving detection accuracy. The accuracy of the proposed model was 0.6 at the beginning. After 300 iterations, the accuracy was 0.85. After 500 iterations, the highest accuracy was close to 0.9. The mAP of the improved model on the dataset reached 90.5%, which was higher than the 78.2% of the base model. The recall rate was 92.0%, the precision was 89.5%, and the calculation time was 220 ms. Meanwhile, the model shows balanced performance in detecting movements of different types of sports, especially in recognizing complex movements such as gymnastics and diving. This model effectively improves the efficiency and accuracy of time action detection through the collaborative action of multiple modules, demonstrating good applicability and robustness.

Abstract Traditional auditing methods have difficulties in detecting various financial issues hidden in massive amounts of data. With the continuous advancement of deep learning and digital technology, new audit methods have been provided for computer auditing. Therefore, in order to achieve intelligent analysis and processing of audit reports, this study innovatively applies the convolution operation of convolutional neural networks to the forward and backward layers of bidirectional long short-term memory networks, obtaining more accurate feature recognition and prediction results. Meanwhile, by introducing radial basis functions for nonlinear mapping of the data space, the model’s ability to fit complex data is enhanced, thereby improving the analytical capability of the digital intelligent financial audit system. The experimental results show that the comprehensive average accuracy of the improved algorithm reaches 92.31%, and the F1-score of the reconciliation function reaches 80.55, which are significantly higher than the other four algorithms. This indicates that the digital intelligent financial audit model proposed in this study can accurately analyze financial audit data, proving that it can comprehensively process various types of financial data and effectively improve the efficiency of modern enterprise audit data analysis.

Abstract Examining the deflection properties while designing and analyzing structural elements offers important information about working of visco-thermoelastic plates in microelectromechanical systems, sensors, and flexible electronic systems. The aim of the current study is to increase the stability and durability in heat-sensitive situations. This study focuses on the transverse deflection in a homogeneous, isotropic, visco-thermoelastic rectangular plate under the influence of uniform loading. The edges of plate are considered at different boundary conditions. Boundary conditions can either involve all edges being simply supported or keeping them simply supported and clamped alternatively. The deflection analysis utilizes the Laplace transform and Finite Fourier sine transform. Inverse Laplace transform is resolved using residue method. MATLAB software aids numerical computation of deflection expression post-Inverse Laplace Transform. Graphical analysis explores the diverse boundary conditions across different modes. The variation in deflection has been studied in the plate under varied circumstances. The findings provide useful insights for engineering applications involving thermal and mechanical stability and advance theoretical and computational understanding of deflection dynamics in visco-thermoelastic rectangular plates under uniform loads.

Abstract The dynamic behavior of the straddle monorail pantograph system is significantly influenced by nonlinearities arising from its unique structural configuration and operational environment. This article presents a mathematical model that incorporates both nonlinear geometric and material properties of the pantograph. A finite element model is utilized to perform modal and static analyses, identifying stress concentrations and dynamic characteristics. To account for random vibration excitations, a nonlinear dynamic response framework is developed, considering vehicle-induced vibrations and network irregularities as stochastic excitation sources. The results demonstrate that these nonlinearities amplify specific resonances under random excitation, leading to increased stress and deformation at critical points. These findings provide a foundation for enhancing fault detection strategies and designing more robust pantographs, with nonlinear vibration monitoring proposed as an effective diagnostic tool. This study contributes to both theoretical understanding and practical improvements in ensuring the reliability and longevity of straddle monorail systems.

Abstract Kuijpers’s study, published in Science under the title “Cavitation-induced reactions in high-pressure carbon dioxide,” explores the phenomenon of acoustic cavitation in high-pressure liquid CO2. However, an analysis of the study suggests that the vapor pressure within bubbles in high-pressure liquid CO2 cannot remain constant or be balanced by static pressure, challenging the fundamental conditions required for acoustic cavitation. A critical prerequisite for cavitation is that the sound pressure must be proportional to the hydrostatic pressure, a condition that does not hold in Kuijpers’s experiments. This raises questions about the interpretation of the ultrasonic effects reported in the study, suggesting that they may not be caused by cavitation. The controversy surrounding acoustic cavitation in high-pressure CO2 is of significant academic interest, as it has implications for fields such as chemical processing, materials science, and ultrasound-assisted reactions. While the physical properties of supercritical CO2 closely resemble those of liquid CO2, experiments conducted with supercritical CO2 indicate that metal corrosion and polymer formation can occur in the absence of cavitation. Moreover, computational simulations have further demonstrated the mechanical effects of ultrasound in both liquid and supercritical CO2, reinforcing the need for a more precise understanding of the mechanisms involved.