Expert Systems with Applications最新文献

{"title":"3D localization for light-field microscopy via convergent accelerated inertial algorithm","authors":"Jinjia Wang ,&nbsp;Shixue Chen ,&nbsp;Xiaofan Wang ,&nbsp;Zhiyuan Deng ,&nbsp;Changle Wang ,&nbsp;Jing Li","doi":"10.1016/j.eswa.2025.127494","DOIUrl":"10.1016/j.eswa.2025.127494","url":null,"abstract":"<div><div>Light-field microscopy (LFM) enables the rapid, light-efficient, and volumetric imaging of target samples within a three-dimensional (3D) scene. The capture of depth and phase information provides insight into the 3D morphology, which is of particular interest in the field of neuroscience. This paper presents an enhanced version of an existing 3D localization method for light-field microscopy offering a notable improvement in both localization accuracy and efficiency. To address the issue of patch overlap, a slice-based convolutional sparse coding problem with a synthesized depth-correlated dictionary is proposed for 3D localization. As the ADMM algorithm does not guarantee convergence, a convergent Nesterov’s accelerated inertial proximal gradient with dry friction (NIPGDF) algorithm is proposed. Furthermore, we have augmented NIPGDF with asymptotic vanishing damping, thereby facilitating rapid convergence. The experimental results demonstrate that NIPGDF offers rapid convergence and high accuracy, effectively detecting the 3D location of target samples in both ideal and noisy environments, and outperforming the ADMM algorithm. Additionally, NIPGDF has been demonstrated to be an effective approach for tracking moving targets in 3D localization experiments with dynamic point sources in a simulated microfluidic environment. This provides a basis for further research into the task of dynamic target tracking. The code is available.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"291 ","pages":"Article 127494"},"PeriodicalIF":7.5,"publicationDate":"2025-06-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144255320","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"AD-RRT*: An RRT*-based global path planning approach for underwater gliders with alpha shapes and DBSCAN","authors":"Yang Li,&nbsp;Rongshun Juan,&nbsp;Yatao Zhou,&nbsp;Tianshu Wang,&nbsp;Zezhong Li,&nbsp;Wei Guo,&nbsp;Zhongke Gao","doi":"10.1016/j.eswa.2025.128219","DOIUrl":"10.1016/j.eswa.2025.128219","url":null,"abstract":"<div><div>In the past decade, Rapidly-exploring Random Tree star (RRT*) and its extensions have been widely applied in robotic path planning due to their asymptotic optimality. This paper propose a novel global path planning method for underwater gliders, called the Alpha shapes and Density-Based Spatial Clustering of Applications with Noise (DBSCAN)-based Rapidly-exploring Random Tree star (AD-RRT*). In this framework, on the basis of considering ocean currents conditions as well as the start and goal points, alpha shapes and DBSCAN are utilized to construct a preferred sampling strategy. In addition, we propose a feasibility assessment to ensure the validity of node connections. Building on this, a circular region sampling strategy inspired by the Monte Carlo method is proposed to enhance overall planning efficiency while maintaining feasibility. To further enhance the exploration process in ocean environments, we propose an ocean currents influence metric to guide parent node selection. Subsequently, edges are rewired based on the estimated travel time, and a time-based iterative optimization framework is employed to optimize the planned paths. Together, these three enhancements significantly improve the efficiency and adaptability of path planning. Finally, simulation experiments demonstrate the superiority of the proposed AD-RRT* method over related approaches, as well as the indispensable role of key components within the overall framework. Future work will focus on local path planning and combining both aspects to enhance the overall path planning of underwater gliders.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"291 ","pages":"Article 128219"},"PeriodicalIF":7.5,"publicationDate":"2025-06-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144255321","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"TopoDiff: Training-free image generation with topological layout control","authors":"Shitong Cao,&nbsp;Xuejie Zhang,&nbsp;Jin Wang,&nbsp;Xiaobing Zhou","doi":"10.1016/j.eswa.2025.128556","DOIUrl":"10.1016/j.eswa.2025.128556","url":null,"abstract":"<div><div>Recent diffusion models can generate high-quality images from text, but their spatial control remains limited. To address this, the goal is to enhance layout control in text-to-image generation without requiring retraining of existing models. Specifically, the proposed TopoDiff framework is a training-free approach that leverages topological guidance to enable precise spatial control during inference. It leaves the original architecture and parameters of Stable Diffusion unmodified. This approach employs a graph-based topological language to explicitly capture object spatial relationships while integrating topological loss into the diffusion model’s denoising process. Additionally, a dynamic offset mechanism is designed to adjust spatial positions during generation, balancing topological structure consistency with the flexibility required for complex generation. Experimental results demonstrate that TopoDiff achieves over 10 % higher Average Precision (AP) than the Stable Diffusion. The source codes are publicly available at <span><span>https://github.com/marcocst/TopoDiff</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"291 ","pages":"Article 128556"},"PeriodicalIF":7.5,"publicationDate":"2025-06-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144280407","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Short-term load forecasting in smart grids: A CGAN-self data reconstruction and BiTCN-BiGRU-self attention model with demand response optimization","authors":"Jingzheng Li,&nbsp;Zhiwen Zhao,&nbsp;Tao Jin","doi":"10.1016/j.eswa.2025.128553","DOIUrl":"10.1016/j.eswa.2025.128553","url":null,"abstract":"<div><div>With the rapid development of smart grids, accurate load prediction is essential for stable operation and optimal scheduling. This paper addresses errors, missing values, and anomalies in electrical load data using a conditional generative adversarial network (CGAN) with dual self-attention (SELF) for data reconstruction. The model simplifies time-series complexity and historical load patterns, eliminating the need for intricate spatiotemporal modeling. Based on the reconstructed data, a short-term load forecasting method is proposed using a bidirectional temporal convolutional network (BiTCN), bidirectional gated recurrent unit (BiGRU), and self-attention. This model processes forward and backward time-series information in parallel, extracting multi-scale features for more accurate predictions. In order to accurately describe the response behavior of users under different electricity price differentials, a logistic demand response (DR) model considering time lag factors is introduced. The model defines optimistic and pessimistic response curves, effectively reflecting the actual range of user responses to price incentives, thus enhancing the practicality of load forecasting in decision support. Experimental results demonstrate that the proposed method not only enhances the accuracy and stability of load forecasting but also provides robust technical support for the stable operation of smart grids.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"292 ","pages":"Article 128553"},"PeriodicalIF":7.5,"publicationDate":"2025-06-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144307110","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"FWLMkNN: Efficient functional K-nearest neighbor based on clustering and functional data analysis","authors":"Mohammed Sabri,&nbsp;Rosanna Verde,&nbsp;Antonio Balzanella","doi":"10.1016/j.eswa.2025.128567","DOIUrl":"10.1016/j.eswa.2025.128567","url":null,"abstract":"<div><div>The increase in data characterized by continuous time and space-varying sequences of observations, such as curves, surfaces, and trajectories, has established the fundamental role of functional data analysis (FDA) in modern statistical methodology. This paper introduces an innovative classification framework that enhances the accuracy of functional data classifiers. This approach merges the strengths of functional supervised and unsupervised learning techniques. It introduces a unique objective function for the unsupervised learning stage to discover novel patterns that are critical for the successful classification of functional data. The process begins with a clustering phase as a preprocessing step that sets the groundwork for the subsequent classification process, which is guided by the clustering results. A partition of the original classes of the training set into distinct subgroups is provided by optimizing a new objective function. This process is achieved by decreasing the variability within each subgroup of a given class while improving the separation between these subgroups and those of other classes. The algorithm automatically determines representative subgroups and the weights assigned to the variables. The weight optimization technique identifies the most discriminative variables for clustering by dynamically adjusting weights to minimize the influence of noise-inducing features in the classification process. Hence this strategy allows for a more efficient and robust classification. Our proposal employs a weighted local mean k-nearest neighbor (KNN) approach within the classification phase. The proposed methodology leverages the novel augmented label space derived from the initial clustering phase, enhancing the classification process. Specifically, the method entails identifying the <span><math><mi>k</mi></math></span> nearest neighbors within each subgroup, computing <span><math><mi>k</mi></math></span> distinct local mean vectors, and subsequently utilizing these vectors to determine their weighted distance relative to the query sample. Consequently, the classification of the query sample is achieved by allocating it to the category exhibiting the minimum distance. The proposed methodology was evaluated using both synthetic datasets and established real-world datasets. Experimental results demonstrate significant reductions in classification error rate compared to state-of-the-art methods, highlighting the framework’s robustness across diverse data. Furthermore, we validate our approach through a practical case study on seasonal classification of Italian electricity load curves, demonstrating its effectiveness in real-world energy management applications.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"292 ","pages":"Article 128567"},"PeriodicalIF":7.5,"publicationDate":"2025-06-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144313924","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An extended prior distribution-based Bayes formula method for cumulative time-dependent failure probability function","authors":"Yingshi Hu ,&nbsp;Zhenzhou Lu ,&nbsp;Jingyu Lei ,&nbsp;Ning Wei ,&nbsp;Jinghan Hu ,&nbsp;Wenhao Li ,&nbsp;Jing Lin","doi":"10.1016/j.eswa.2025.128551","DOIUrl":"10.1016/j.eswa.2025.128551","url":null,"abstract":"<div><div>Since the cumulative time-dependent failure probability function (CTFPF) can provide the time-dependent failure probability (TFP) with respect to distribution parameters and upper bound of time interval (UBTI), estimating CTFPF can provide great convenience for solving time-dependent reliability-based design optimization. However, the existing direct Monte Carlo simulation method (MCS) for estimating CTFPF is time-consuming. Therefore, this paper proposes an extended prior distribution-based Bayes formula method (EPD-Bayes) to improve the efficiency and accuracy of estimating CTFPF. The EPD-Bayes adopts the Bayes formula to transform the focus of estimating CTFPF into efficiently estimating the time-dependent failure domain under different UBTI. Then, a first failure instant (FFI) learning function combined with adaptive candidate sample pool reduction technology (ACSPRT) is established to efficiently obtain the time-dependent failure domain under different UBTI. At the meanwhile, to avoid the boundary effect of kernel density estimation method (KDE) in estimating the conditional probability density function (PDF) of distribution parameters, an extended prior distribution is proposed to improve the accuracy of estimating the conditional PDF at the boundary of distribution parameter space. The results of three examples verify the advantage of the proposed EPD-Bayes.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"291 ","pages":"Article 128551"},"PeriodicalIF":7.5,"publicationDate":"2025-06-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144270026","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"SADiff: Structure-aware diffusion model for enhancing prostate multiphoton microscopy imaging","authors":"Maoye Huang ,&nbsp;Xinpeng Huang ,&nbsp;Hong Chen ,&nbsp;Haoyi Fan ,&nbsp;Peng Shi ,&nbsp;Xiaoqin Zhu","doi":"10.1016/j.eswa.2025.128447","DOIUrl":"10.1016/j.eswa.2025.128447","url":null,"abstract":"<div><div>Multiphoton Microscopy (MPM) has become an essential technique in bioimaging, particularly for studying thick tissues and live animals, offering deep tissue penetration, high-resolution imaging, and minimizing photodamage and photobleaching. Nevertheless, MPM comes with performance trade-offs related to imaging quality, acquisition speed, and sample health in practice. Achieving high-resolution imaging with optimal signal-to-noise ratio (SNR) in a timely manner while minimizing potential sample damage remains challenging, especially when the goal is to provide pathologically relevant diagnostic information in clinical settings. In this paper, we propose a structure-aware diffusion model (SADiff), specifically designed to enhance denoising and super-resolution in prostate MPM imaging. SADiff innovatively incorporates high-frequency residuals into the diffusion process, a strategy that significantly improves the model’s ability to capture and preserve fine structural details, such as glandular morphology, that are crucial for accurate prostate cancer diagnosis, without introducing additional trainable parameters. Moreover, a novel symmetric tanh-based schedule is designed to effectively control the integration of high-frequency residuals, ensuring optimal image quality. Extensive experiments conducted on clinical prostate cancer MPM image datasets demonstrate that SADiff significantly outperforms state-of-the-art methods in both qualitative and quantitative evaluations. By transforming low-resolution, low-SNR images into high-resolution, high-SNR images, SADiff provides a robust solution for MPM imaging, enhancing diagnostic accuracy and potentially leading to improved patient outcomes.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"292 ","pages":"Article 128447"},"PeriodicalIF":7.5,"publicationDate":"2025-06-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144314182","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Decision-analytics-based enhancing auditor performance evaluation: Application of multi-criteria methodology with Monte Carlo simulation","authors":"Ahmet Kaya ,&nbsp;Hasan Emin Gurler ,&nbsp;Nazan Güngör Karyağdı ,&nbsp;Mehmet Özçalıcı ,&nbsp;Yusuf Akpınar ,&nbsp;Nurettin Koca ,&nbsp;Dragan Pamucar","doi":"10.1016/j.eswa.2025.128561","DOIUrl":"10.1016/j.eswa.2025.128561","url":null,"abstract":"<div><div>This study highlights the critical importance of auditor performance, as it directly influences the quality and reliability of financial reporting, which is crucial for maintaining trust in financial markets. High-performing auditing firms contribute to improved transparency and accountability, which are essential for effective corporate governance and investor confidence. The study evaluates the performance of 27 auditing firms in Turkey using two advanced MCDM methods: WENSLO and ARTASI. By focusing on measurable, operational factors such as the number of audited companies, training duration, and client exposure, the study provides a more data-driven and objective approach to performance evaluation. The WENSLO method assigns weights to these criteria, while ARTASI ranks the firms accordingly. According to the WENSLO results, the two most important criteria among those examined are the number of completed audits and the duration of training. The results reveal that PKF, HSY, and YEDITEPE are the top performers, while ECOVIZ, VEZIN, and REFORM ranked the lowest. The study’s contribution lies in its novel approach to performance assessment, offering a comprehensive, data-driven evaluation model that emphasizes quantifiable metrics. It also demonstrates the robustness of the findings through sensitivity analyses and comparisons with other MCDM techniques, such as MABAC and TOPSIS. The study also incorporates a Monte Carlo simulation, which tested the impact of random weight variations on the ARTASI rankings. The simulation confirmed the robustness of the rankings, revealing that certain firms consistently outperformed others, regardless of changes in the weight distribution. The findings suggest that industry specialization and operational factors play a significant role in auditing performance, and future research could expand these criteria or explore additional MCDM methods to enhance the reliability and generalizability of results.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"291 ","pages":"Article 128561"},"PeriodicalIF":7.5,"publicationDate":"2025-06-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144270676","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Efficient congestion-aware routing in MANETs using quantum self-attention and optimized siamese network","authors":"Rajagopal Reka ,&nbsp;Bade Rebecca ,&nbsp;Murugavelu Mathivanan ,&nbsp;Muthusamy Rameshkumar","doi":"10.1016/j.eswa.2025.128560","DOIUrl":"10.1016/j.eswa.2025.128560","url":null,"abstract":"<div><div>Mobile Ad Hoc Networks (MANETs) are dynamic, self-configuring wireless networks that frequently undergo topology changes, making them highly susceptible to link errors, congestion, and excessive energy consumption. Existing routing protocols struggle with real-time congestion detection and efficient path selection, leading to higher packet loss and increased network overhead. To address these challenges, this study proposes an adaptive Quantum Self-Attention-based Triple Pseudo Siamese Network (QSA-TPSN) with the Walrus Optimizer (WO) for congestion-aware and energy-efficient routing in MANETs. The QSA module enhances congestion detection by prioritizing stable, high-quality routes, reducing retransmissions by 25%. The TPSN framework learns congestion patterns and dynamically refines routing decisions to minimize delay. Meanwhile, the WO optimizer optimally selects paths based on real-time congestion and energy metrics, ensuring load balancing and efficient resource utilization. The QSA-TPSN-WO model is evaluated against CLEE, CL-QAERP, ESCL-PSO, and ANN-based routing approaches using packet delivery ratio (PDR), delay, throughput, energy consumption, and packet drop rate as performance metrics. Experimental results confirm that QSA-TPSN-WO achieves a 99.4% PDR, reduces packet loss by 28%, decreases energy consumption by 30%, and improves throughput by 15% compared to state-of-the-art methods. These findings demonstrate that QSA-TPSN-WO significantly enhances routing stability, congestion control, and energy efficiency, making it a robust solution for MANET environments.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"291 ","pages":"Article 128560"},"PeriodicalIF":7.5,"publicationDate":"2025-06-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144280405","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"PRCD: A full-chain parallel residual compensation debiasing framework","authors":"Chenyang Li ,&nbsp;Maoyuan Zhang ,&nbsp;Meng Zheng","doi":"10.1016/j.eswa.2025.128544","DOIUrl":"10.1016/j.eswa.2025.128544","url":null,"abstract":"<div><div>Hate speech detection models often suffer from systemic misclassification due to biases in training data, particularly cognitive distortions in semantic associations learned in deep networks, making it especially challenging to identify implicit biases. While existing neuron pruning methods can mitigate explicit biases to some extent, removing parameters weakens the model’s semantic representation capability and struggles to address deeply ingrained cognitively distorted associations in deep networks. To tackle this issue, this paper proposes a full-chain parallel residual compensation debiasing framework (PRCD). This framework introduces a residual compensation method for soft pruning (RCM), which enables soft pruning of bias signals across the entire model, from shallow to deep layers—without compromising the model’s semantic representation ability. Additionally, a toxicity constraint enhancement method based on sensitivity attribution prediction (ToxCon) is incorporated to generate contrastive samples that expose bias, effectively guiding RCM in correcting implicit biases stemming from cognitive distortions in semantic associations. Experimental results on three public datasets demonstrate that the PRCD significantly improves model performance and fairness in detecting hate speech, achieving state-of-the-art performance.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"291 ","pages":"Article 128544"},"PeriodicalIF":7.5,"publicationDate":"2025-06-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144280501","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}