Expert Systems with Applications最新文献

{"title":"Discrete zeroing neural dynamic with noise tolerance for image deblurring","authors":"Cong Lin ,&nbsp;Fenghao Zhuang ,&nbsp;Jiahao Li ,&nbsp;Chengze Jiang ,&nbsp;Yuanyuan Wu","doi":"10.1016/j.eswa.2025.128914","DOIUrl":"10.1016/j.eswa.2025.128914","url":null,"abstract":"<div><div>As the demand for high-quality images continues to grow, image deblurring has become a fundamental challenge in computer vision. Although numerous effective deblurring methods have been proposed, one critical area remains largely unexplored: the interference caused by environmental noise. It is well known that noise can perturb solution systems, leading to instability or even collapse. Current mainstream methods, such as deep learning-based approaches, struggle to address such perturbations effectively. Additionally, these methods require large datasets for training and optimization, which incur significant computational cost and time. Without sufficient data, their robustness and deblurring performance are greatly limited. To address these challenges, we consider an alternative approach: a learning-free neural network, called neural dynamic. Our method employs a dynamic solving mechanism capable of addressing potential static optimization problems, while its integral term enhances noise resistance. To further adapt this framework for practical engineering applications, we developed a Taylor expansion-based discretization scheme called Taylor-type 6-instant Noise-Tolerance Zeroing neural Dynamic (T6NTZD). This model not only improves noise resistance but also achieves lightweight design and real-time processing. By introducing this approach, we aim to fill a significant gap in the field of image deblurring. Finally, through a detailed theoretical analysis from a continuous perspective and a comprehensive comparison with 12 neural dynamics models, the superiority of this method is clearly demonstrated. The key advantages of our model are summarized as follows: strong robustness, lightweight design, and the elimination of the need for data-intensive learning.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"296 ","pages":"Article 128914"},"PeriodicalIF":7.5,"publicationDate":"2025-07-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144588208","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":"Dynamic sanctioning mechanism for cooperative multi-agent systems","authors":"Linjie Liu ,&nbsp;Lichen Wang ,&nbsp;Weiyan Niu ,&nbsp;Shijia Hua","doi":"10.1016/j.eswa.2025.128873","DOIUrl":"10.1016/j.eswa.2025.128873","url":null,"abstract":"<div><div>Coordinating multi-agent systems to accomplish complex tasks presents a profound and unprecedented challenge, introducing significant uncertainty into the operational framework of collective artificial intelligence. Addressing this formidable challenge requires collective actions of cooperation and concerted efforts on a global scale. However, progress in addressing this challenge through voluntary contributions has been regrettably slow, highlighting the need for the participation of robust sanctioning mechanisms to drive meaningful change. Here, we propose a dynamic sanctioning framework that relies on adjusting between positive and negative incentives based on the collective status of the population. We show that the transition of sanctioning institutions from punitive measures to rewarding mechanisms can effectively sustain a high level of cooperation, even when the risk of collective action failure is low. The threshold at which sanctioning institutions choose to switch incentives plays a crucial role in shaping evolutionary outcomes. Moreover, we provide further evidence that the success of the reward mechanism is based on the presence of self-interested altruism, in which the implementing authority also benefits from the incentives.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"296 ","pages":"Article 128873"},"PeriodicalIF":7.5,"publicationDate":"2025-07-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144588760","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":"Knowledge-embedded graph representation learning for document-level relation extraction","authors":"Jinglin Liang ,&nbsp;Yutao Qin ,&nbsp;Shuangping Huang ,&nbsp;Yunqing Hu ,&nbsp;Xinwu Liu ,&nbsp;Huiyuan Zhang ,&nbsp;Tianshui Chen","doi":"10.1016/j.eswa.2025.128872","DOIUrl":"10.1016/j.eswa.2025.128872","url":null,"abstract":"<div><div>Document-level relation extraction (DocRE) aims to identify the relations between entities in a document, which serves as a fundamental task in natural language processing. In DocRE, there inherently exists strong prior knowledge, for example, individuals and organizations tend to exhibit a “membership” relation rather than a “separation” relation. Leveraging this knowledge can effectively confine the model’s prediction space and highlight potential relations between entities. However, existing DocRE works primarily focus on designing sophisticated models to implicitly encode document features, inadvertently neglecting this informative prior knowledge, which might lead to suboptimal performance. In this work, we assume that the prior knowledge can be effectively represented by statistical co-occurrence correlations between entity types and relations. Based on this premise, we propose a novel algorithm called Knowledge-Embedded Graph Representation Learning (KEGRL), which enhances the representation of entity features through this statistical prior knowledge. Specifically, we calculate the statistical co-occurrence correlations existing between entity types and relations. These correlations are then ingeniously encapsulated within weighted edge-oriented heterogeneous graphs, where nodes correspond to entities and relations. Every entity node is connected to all relation nodes, and their edges symbolize the statistical correlations between entities and relations. Across these graphs, the entity features propagate under the guidance of statistical correlations, during which the statistical knowledge is injected into the entity features to enhance their distinctiveness. Extensive experiments on multiple baseline models and datasets consistently demonstrate that integrating KEGRL significantly enhances the performance of DocRE models. The code is available at <span><span>https://github.com/MrDreamQ/KEGRL</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"295 ","pages":"Article 128872"},"PeriodicalIF":7.5,"publicationDate":"2025-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144588260","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":"Simplifying the operation of hybrid ad-hoc sensor networks with neural networks as the sole data reconstruction method","authors":"Piotr Cofta","doi":"10.1016/j.eswa.2025.128981","DOIUrl":"10.1016/j.eswa.2025.128981","url":null,"abstract":"<div><div>Hybrid <em>ad hoc</em> sensor networks, which are known for their cost-effective sensors and opportunistic yet affordable management, are becoming increasingly popular. In such networks, in addition to data reconstruction, various different methods are used to address operational problems, such as failure of sensors, resilience against attacks, loss of calibration, etc. A combination of several methods may not always be practical or optimal. However, this research demonstrates that it is not only possible but also beneficial to use a single data reconstruction method instead, thus decreasing the operational cost. Artificial neural networks, and particularly the multi-layer perceptron, are proposed as a single method. Through simulations, nine scenarios are analyzed to demonstrate the fitness for purpose of this approach. The findings demonstrate that the multi-layer perceptron can not only be used as a sole data reconstruction method, but also consistently improves the quality of data reconstruction across all the scenarios tested here.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"296 ","pages":"Article 128981"},"PeriodicalIF":7.5,"publicationDate":"2025-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144605718","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 AI-optimized strategy for intelligent risk mapping of Islamic and conventional sustainable markets: Assessing the enduring dynamics of technological risk spillovers","authors":"Mahdi Ghaemi Asl","doi":"10.1016/j.eswa.2025.128945","DOIUrl":"10.1016/j.eswa.2025.128945","url":null,"abstract":"<div><div>This study explores the lasting impact of industries influenced by Robotic-Artificial Intelligence-Cloud (RAIC) technologies on risk management in both conventional and Islamic sustainable markets, employing a novel AI-driven framework. By utilizing the Quantile-based Total Connectedness Index (QTCI) to gauge market interconnectedness and Long Short-Term Memory (LSTM) neural networks to evaluate risk persistence, the research investigates how sectors such as autonomous vehicles, cybersecurity, cleantech, and future payments influence financial stability across different market conditions (bull, bear, and normal). The findings reveal divergent risk dynamics: Islamic markets are more sensitive to technological disruptions, particularly in robotics and cybersecurity, while conventional markets show more stable integration with sectors like smart grids and space technologies. Cleantech shows a tendency to coincide with decreased market volatility during bear markets, while future payments demonstrate widespread interconnectedness across all market conditions. AI-driven analysis highlights those Islamic markets excel in risk mitigation during stable conditions but conventional markets exhibit greater adaptability in the face of change. The QTCI-LSTM hybrid approach identifies differences in risk persistence, showing that technologies like genetic engineering and nanotechnology have transient effects in Islamic markets but more enduring roles in conventional markets. The study offers policy recommendations for sector-specific strategies, advocating for enhanced resilience in volatile sectors during bull markets, prioritizing cleantech during downturns, and encouraging cross-market collaboration. This work contributes to sustainable finance literature by integrating AI-powered persistence analysis with traditional risk metrics. The findings offer insights for policymakers managing technological integration in evolving markets.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"296 ","pages":"Article 128945"},"PeriodicalIF":7.5,"publicationDate":"2025-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144605062","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":"DPTrack: Dual-prompt guided network for visual object tracking","authors":"Kang Liu,&nbsp;Long Liu,&nbsp;Jiaqi Wang,&nbsp;Pingyan Hu,&nbsp;Yunhe Wang","doi":"10.1016/j.eswa.2025.128974","DOIUrl":"10.1016/j.eswa.2025.128974","url":null,"abstract":"<div><div>Recently, prompt-based spatio-temporal trackers have achieved impressive advancements. However, these methods tend to incorporate only limited spatio-temporal information through prompts, such as recursive prompts or historical prompts, which fail to fully exploit spatio-temporal context information in video sequences, i.e., the instantaneous states of nearby frame and the consistent states of past frames. This oversight inhibits further performance improvement in complex scenes. To tackle this issue, we resort to prompt learning and present a dual-prompt guided visual tracking network (DPTrack), consisting of an instantaneous prompt network (IPN) and a spatio-temporal consistency prompt network (ST-CPN). Specifically, the IPN captures the appearance changes of the nearby frame to generate instantaneous prompt, which are directly involved in feature extraction and information interaction. The ST-CPN learns a set of learnable prompts to summarize the spatio-temporal consistency of previous frames, and then iteratively guides the search features to emphasize the target embedding. In this way, the dual-prompt exploits rich spatio-temporal cues, enhancing the adaptability and robustness of the tracker. Furthermore, we introduce a spatio-temporal pooling collection mechanism (SPC) to maintain consistency and adapt to the appearance changes. Extensive experiments on seven benchmarks prove that the proposed DPTrack achieves very promising tracking performance. Our DPTrack achieves 77.6 % AO on GOT-10k and 52.8 % AUC on LaSOT<span><math><msub><mrow></mrow><mrow><mi>e</mi><mi>x</mi><mi>t</mi></mrow></msub></math></span>. Notably, it obtains 73.9 % AUC, 81.8 % P, and 84.5 % P<span><math><msub><mrow></mrow><mrow><mi>N</mi><mi>o</mi><mi>r</mi><mi>m</mi></mrow></msub></math></span> on LaSOT, 60.8 % EAO, 78.2 % A, and 89.3 % R on VOT2020, outperforming the remarkable trackers and demonstrating its superiority.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"296 ","pages":"Article 128974"},"PeriodicalIF":7.5,"publicationDate":"2025-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144631518","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":"High-fidelity mural inpainting via progressive reconstruction and damage-aware adaptation","authors":"Shuyi Qu ,&nbsp;Qingqing Kang ,&nbsp;Zhe Yu ,&nbsp;Shenglin Peng ,&nbsp;Jun Wang ,&nbsp;Qiyao Hu ,&nbsp;Xianlin Peng ,&nbsp;Jinye Peng","doi":"10.1016/j.eswa.2025.128957","DOIUrl":"10.1016/j.eswa.2025.128957","url":null,"abstract":"<div><div>Computer vision techniques have revolutionized digital mural inpainting. However, single-stage networks often yield suboptimal results with blurred textures and structural distortion, while existing progressive strategies struggle to effectively balance local and global information. To address these limitations, we propose a novel generative adversarial model that progressively reconstructs mural details by adaptively integrating multi-scale local features and global context based on damage severity. We first obtain initial coarse results using an encoder-decoder network. Then, a mask-guided network adaptively extracts and fuses local features according to damage levels. Next, multi-level residual learning further refines details at different scales. Finally, a global network captures overall artistic characteristics using an optimized Transformer-UNet architecture. In this way, our method harmonizes detailed local restoration with the preservation of overall artistic integrity throughout the progressive inpainting process. Extensive experiments on multiple mural datasets demonstrate that our method achieves state-of-the-art performance in terms of texture clarity and structural coherence. We release the source code at <span><span>https://github.com/Kk01Qq/Mural-Inpainting</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"296 ","pages":"Article 128957"},"PeriodicalIF":7.5,"publicationDate":"2025-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144613829","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":"A modified domain adversarial approach based on model and data-driven for bearing fault diagnosis","authors":"Ning Zhang ,&nbsp;Zhaohui Qiao ,&nbsp;Baosu Guo ,&nbsp;Fenghe Wu ,&nbsp;Junwei Fan","doi":"10.1016/j.eswa.2025.128970","DOIUrl":"10.1016/j.eswa.2025.128970","url":null,"abstract":"<div><div>Bearing fault diagnosis is essential for maintaining the reliability and stability of mechanical equipment. However, obtaining sufficient labeled data in real scenarios is high-risk and challenging, which limits the further application of traditional data-driven approaches. In this research, a novel model and data-driven approach called the modified domain adversarial neural network (MDANN) is developed for bearing fault identification. Specifically, a bearing dynamic model is established, so that the priori information on bearing failures can be acquired by finite element simulation. A data-driven MDANN model is then developed for feature extraction and cross domain transfer from simulated data to measured data. The attention module is introduced for feature weight reassignment, so that the priority of domain-invariant features is increased and negative transfer is suppressed. The improved loss function incorporating adaptive CORAL is designed to align both marginal and conditional distributions. Finally, the validity of the proposed MDANN is validated through two cases. The results demonstrate that the domain transfer capability of MDANN outperforms other methods in cross-domain tasks.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"296 ","pages":"Article 128970"},"PeriodicalIF":7.5,"publicationDate":"2025-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144613853","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 ILS-VND approach to dynamic pricing of perishable products","authors":"Gabriel Villa ,&nbsp;B. Adenso-Díaz ,&nbsp;S. Lozano","doi":"10.1016/j.eswa.2025.128949","DOIUrl":"10.1016/j.eswa.2025.128949","url":null,"abstract":"<div><div>This paper deals with the problem of setting prices of a perishable product whose demand decreases over time due to its perishable character and its price elasticity. It considers a discrete-time, deterministic model whose decision variables are the order quantity and the dynamic pricing policy (modelled as a series of discrete discounts in specific periods). Given the combinatorial structure of the problem and its non-linear nature, a hybrid Iterated Local Search (ILS) + Variable Neighborhood Descent (VND) metaheuristic approach is proposed. The initial solution for the search is computed using a heuristic which generally finds a good starting solution. The proposed approach is rather flexible and can accommodate many different scenarios. In particular, it has been validated on two scenarios: one involving a two-day horizon, with 1 h unit time and 12 h open/12 h closed cycle, and another one that considers a 48-day horizon, with 6 h unit time and 12/6 h open cycle. The results show that, in both cases, the proposed metaheuristic outperforms Simulated Annealing (SA), achieves a slight improvement over the heuristic, and reaches the optimal solution (verified through complete enumeration) while maintaining low computational costs. It has also been shown that profit increases of almost 20 %, compared to the no-discount policy.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"295 ","pages":"Article 128949"},"PeriodicalIF":7.5,"publicationDate":"2025-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144580107","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":"EEG-DTIL: An EEG-based dynamic task-incremental learning method for decoding ADL-oriented motor imagery pairs","authors":"Yufei Yang ,&nbsp;Mingai Li ,&nbsp;Fubiao Huang","doi":"10.1016/j.eswa.2025.128927","DOIUrl":"10.1016/j.eswa.2025.128927","url":null,"abstract":"<div><div>Patients with strokes are likely to suffer from dyskinesia and lose their ability to perform activities of daily living (ADL). Motor neurorehabilitation can be gradually realized by continuously learning ADL-oriented motor imagery (MI) pairs with opposite directions, and an electroencephalography (EEG)-based brain-computer interface (BCI) is an effective solution. However, decoding the undetermined streams of MI pairs remains a great challenge for maintaining the balance between new and old tasks. Thus, an EEG-based dynamic task-incremental learning method, called EEG-DTIL, is proposed for decoding progressively incoming MI pairs. Based on the wavelet packet transform, all of the reconstructed subband signals are applied to extract global-view spatial features (GSF) via the regularized common spatial pattern, and the preferred parts are used to capture local-view spatial features (LSF) via tangent space mapping from the Riemannian space. Multiview spatial features (MvSF) are obtained after performing fusion and dimensionality reduction on the GSF and LSF. Inspired by the broad learning system (BLS), a series of personalized sub-BLSs are created in the same order as the inflowing MI pairs and used to construct a residual-based stacked BLS (R-SBLS). Moreover, a dynamic weight consolidation block (DWC) is developed to remember more learned knowledge by controlling the key output weights to be updated within a low error range. Finally, R-SBLS and DWC are combined in parallel, forming a dynamic incremental learning network (DRI-Net). On public and self-collected datasets, EEG-DTIL achieves incremental decoding accuracies of 70.53% and 71.80% for task streams with two and three MI pairs, respectively. The experimental results demonstrate that EEG-DTIL is significantly superior to the related methods, exhibiting better plasticity for new MI pairs, retainability for old MI pairs, and robustness to MI pair sequences.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"296 ","pages":"Article 128927"},"PeriodicalIF":7.5,"publicationDate":"2025-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144633495","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}