Expert Systems with Applications最新文献

{"title":"A network selection algorithm for space-air-ground integrated network based on location prediction model and multi-attribute decision making","authors":"Jianli Xie,&nbsp;Weicheng Pan,&nbsp;Lei Wu,&nbsp;Zishan Wu","doi":"10.1016/j.eswa.2025.129646","DOIUrl":"10.1016/j.eswa.2025.129646","url":null,"abstract":"<div><div>As an indispensable component of 5G and even the future 6G networks, the Space-Air-Ground Integrated Network (SAGIN) is envisioned to provide ubiquitous network connectivity and services by integrating satellite, aerial, and terrestrial networks. However, due to the frequent network selection of in-vehicle terminals, the user’s Quality of Service (QoS) can significantly deteriorate. To address this issue, a network selection algorithm based on terminal location prediction has been proposed. Firstly, we enhanced the Particle Swarm Optimization (PSO) algorithm to optimize the hyper-parameters of the Long Short-Term Memory (LSTM) network, thereby improving the accuracy of terminal location prediction. After constructing the network sets of the current terminal position and the predicted position, respectively, we designed a network selection judgment mechanism with a dynamically adjustable switching threshold based on Fuzzy Logic and K-Means theory. Finally, through the Technique for Order Preference by Similarity to an Ideal Solution (TOPSIS) algorithm, we have achieved robust network selection in fast-moving scenarios. The simulation results show that the proposed algorithm can adaptively adjust the switching threshold and provide precise positions. Compared to existing algorithms, it can significantly reduce the number of candidate networks and the number of selections, thereby reducing the computational load and increasing the throughput of users.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"298 ","pages":"Article 129646"},"PeriodicalIF":7.5,"publicationDate":"2025-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145158639","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 physics-informed neural network surrogate model and many-objective optimization algorithm for coupled multi-energy systems in smart grids","authors":"Jingbo Zhang ,&nbsp;Xingjuan Cai ,&nbsp;Zhihua Cui ,&nbsp;Jinjun Chen","doi":"10.1016/j.eswa.2025.129760","DOIUrl":"10.1016/j.eswa.2025.129760","url":null,"abstract":"<div><div>The scope of smart grids is progressively extending toward Integrated Energy Systems (IES) that couple electricity with gas, heating, and cooling. Due to the unsteady-state physical characteristics inherent in the transmission of gas, heat, and cooling resources, IES scheduling must not only balance multiple typical objectives but also account for the dynamic coupling of heterogeneous physical domains. To address these challenges, this paper formulates a Many-objective Optimization Model for Coupled Multi-Energy Flows (MaOCMFM) with partial differential equations (PDEs) in IES, which captures the dynamic physical behaviors of electricity, gas, heat, and cooling subsystems. Building upon this model, we propose a Probabilistic Contributing Many-objective Evolutionary Algorithm enhanced by a Physics-Informed Neural Network surrogate model (PC-MaOEA-PINN). Cubic B-spline functions are employed to achieve a continuous representation of the decision variables, while multi-physics constraints are embedded into the loss function of the surrogate model. This design enables efficient approximation of the objective function with a limited number of samples and facilitates focused exploration in critical evolutionary regions, thereby accelerating population convergence. The effectiveness of the proposed model and algorithm is validated on 9 typical scheduling days across four simulated IES scenarios.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"298 ","pages":"Article 129760"},"PeriodicalIF":7.5,"publicationDate":"2025-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145159219","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":"Sustainable time-dependent intermodal hub-and-spoke logistic network considering hub failure: A mathematical model and a hybrid artificial bee colony algorithm","authors":"Burcu Tokbay Erkek ,&nbsp;Salih Himmetoğlu ,&nbsp;Yılmaz Delice ,&nbsp;Emel Kızılkaya Aydoğan","doi":"10.1016/j.eswa.2025.129804","DOIUrl":"10.1016/j.eswa.2025.129804","url":null,"abstract":"<div><div>This paper addresses the design of a sustainable hub-and-spoke logistics network that integrates intermodal transportation between the hubs, hub failures, time dependency, and environmental parameters. Accordingly, we propose a novel mixed-integer linear programming (MILP) model and a hybrid artificial bee colony-based algorithm (HABCb) to minimize transportation costs and emissions in robust network configurations. The model is the first to simultaneously integrate intermodality, sustainability metrics, and hub disruption scenarios within a single framework. Computational experiments using real-life data from Turkey demonstrate that the proposed HABCb approach outperforms both genetic algorithm (GA) and artificial bee colony (ABC) algorithm. On medium-sized problem sets, it achieves average cost reductions of 7% compared to GA and 10% compared to ABC algorithm, while on large-sized problems the reductions are 10% and 15%, respectively. Furthermore, the HABCb approach provides faster convergence and higher-quality solutions for larger problem sizes. The findings highlight the practical and theoretical insights of incorporating sustainability, intermodality, and robustness into hub-and-spoke network design.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"298 ","pages":"Article 129804"},"PeriodicalIF":7.5,"publicationDate":"2025-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145159094","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 hybrid risk assessment method combining CatBoost and FAHP-Grid search optimized risk matrix for container ship accident","authors":"Yuqing Xiao ,&nbsp;Shilian Han ,&nbsp;Xinwang Liu","doi":"10.1016/j.eswa.2025.129763","DOIUrl":"10.1016/j.eswa.2025.129763","url":null,"abstract":"<div><div>As a dominant mode of maritime transportation with unique risk characteristics, container shipping requires accurate and applicable risk assessment. However, conventional risk matrices oversimplify complex interactions, while pure data-driven models lack operational utility. To address this, a hybrid method for container ship risk assessment is proposed. This method integrates CatBoost-based predictive method, FAHP-grid search optimized risk matrix, and GIS-supported risk mapping. A comprehensive study of maritime casualties and piracy accidents is conducted, utilizing historical incident data sets collected from the Global Integrated Shipping Information System (GISIS). The global maritime accident risk of container ships is then evaluated and mapped. The sensitivity analysis confirms the robustness of the method under varying linguistic distance parameters, while expert weights have a moderate impact on the assessment results. Finally, the effectiveness of the proposed method is validated through comparative analyses on predictive performance, risk discrimination capability, and risk assessment accuracy. CatBoost algorithm outperforms XGBoost, LightGBM, and Random Forest algorithms in predictive metrics. The designed risk matrix shows strong discriminatory ability for container ship risk levels. In historical accident data validation, the proposed method also achieves higher accuracy than combinations involving XGBoost, LightGBM, or Random Forest with the designed risk matrix.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"298 ","pages":"Article 129763"},"PeriodicalIF":7.5,"publicationDate":"2025-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145159095","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 novel grey possibility clustering method based on inverse perspective and its applications","authors":"Junjie Wang ,&nbsp;Xun Li ,&nbsp;Yaoguo Dang ,&nbsp;Zhongju Shang ,&nbsp;Li Ye ,&nbsp;Sifeng Liu","doi":"10.1016/j.eswa.2025.129837","DOIUrl":"10.1016/j.eswa.2025.129837","url":null,"abstract":"<div><div>The center-point mixed possibility functions (CMPFs), which are determined by the decision makers qualitatively, are the most important element to obtain the effective clustering results in a grey clustering model. However, different decision makers provide different CMPFs which may lead to inconsistent or contradictory clustering results. In response to this problem, this article proposes an inverse grey possibility clustering model which can determine the CMPFs based on the part of the given final results. This novel matrix-based method can derive all of the required CMPFs which satisfy the partially known clustering results. More specifically, four theorems are put forward to analyze the four different cases, which are single index single object (SISO), single index multiple objects (SIMO), multiple indices single object (MISO) and multiple indices multiple objects (MIMO), respectively, to derive the required CMPFs of a given clustering result using algebraic expressions. For the purpose of developing the matrix representations for the MISO and MIMO situations, a new unified expression of the CMPFs to replace their existing segmented function expression is proposed. Finally, in order to demonstrate how it can be used in practice, the proposed method is applied for evaluating the effects of the reduction of pollution and carbon emissions and determining aerospace equipment component suppliers with different types of data. Compared to the forward GPC models, the proposed IGPC model has higher accuracy.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"298 ","pages":"Article 129837"},"PeriodicalIF":7.5,"publicationDate":"2025-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145220910","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":"Attention-enhanced 3D residual networks for knee abnormality classification","authors":"Mohamad M.A. Ashames ,&nbsp;Semih Ergin ,&nbsp;Omer N. Gerek ,&nbsp;H. Serhan Yavuz","doi":"10.1016/j.eswa.2025.129858","DOIUrl":"10.1016/j.eswa.2025.129858","url":null,"abstract":"<div><div>The advancement of deep learning technologies, particularly through Convolutional Neural Networks (CNNs), has substantially enriched medical image analysis. This study focuses on improving knee MRI diagnostics by comparing 2D and 3D CNN architectures using the MRNet and SKM-TEA datasets. Initially, modified 2D CNNs, such as ResNet50, were applied for plane-specific and integrated multi-plane analyses. Plane-specific models captured detailed anatomical features, while integrated approaches synthesized information across multiple planes, improving diagnostic capability but lacking full volumetric data utilization. To address these limitations, a novel 3D CNN architecture enhanced with residual attention blocks was developed, leveraging volumetric MRI data. These blocks integrate spatial attention and Squeeze-and-Excitation (SE) mechanisms, optimizing feature focus for accurate diagnostics. This approach improved both model precision and interpretability, which are crucial for clinical applications. Experimental evaluation on the MRNet dataset demonstrated that the proposed 3D CNN outperformed 2D models, achieving 83.58 % accuracy for abnormalities. On the SKM-TEA dataset, the model classified Meniscal Tear (71.36 %), Ligament Tear (79.84 %), Cartilage Lesion (84.28 %), and Effusion (76.74 %), demonstrating robustness in complex pathology detection. Gradient-weighted Class Activation Mapping (Grad-CAM) further enhanced interpretability by highlighting critical diagnostic regions. These findings emphasize the effectiveness of attention-guided 3D CNNs in knee abnormality classification. Future work will explore broader applications in medical imaging, refining the model’s generalizability across diverse clinical datasets.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"298 ","pages":"Article 129858"},"PeriodicalIF":7.5,"publicationDate":"2025-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145222080","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":"Enhancing mixture-of-experts model with prior knowledge for infrared and visible image fusion in complex degraded environments","authors":"Gang Li ,&nbsp;Chengrun Jiang ,&nbsp;Jiachen Li ,&nbsp;Jin Wan ,&nbsp;Mingle Zhou ,&nbsp;Delong Han","doi":"10.1016/j.eswa.2025.129844","DOIUrl":"10.1016/j.eswa.2025.129844","url":null,"abstract":"<div><div>Infrared and visible image fusion aims to generate a composite image that simultaneously preserves thermal radiation information from infrared images and the rich texture details of visible images. However, existing studies have overlooked the adverse effects of scene degradation in visible images on the fusion process, leading to suboptimal fusion outcomes. To address the challenges posed by scene degradation in image fusion tasks, this paper proposes an image fusion network with degradation correction capability named the Enhancing Mixture-of-Experts model with Prior knowledge for infrared and visible image fusion (EMPFusion), which pioneers the automated execution of multiple degradation restoration tasks during the fusion process. First, we develop a diffusion model for degradation removal to generate high-quality pseudo-labels of visible images, thereby providing supervisory signals for training the fusion network. Second, to overcome the significant challenges in feature extraction caused by complex and diverse degradation scenarios, we design a Degradation removal backbone based on Prior knowledge and the Mixture-of-Experts (DPM) module. This architecture removes degradation with low loss and moderate computational overhead by integrating domain-specific prior knowledge and the Mixture-of-Experts framework. Furthermore, to mitigate semantic loss under extreme environmental conditions, we propose a Semantic Deconstruction and Segmentation (SDS) module based on image-text foundation models, enhancing semantic consistency throughout the fusion process. Extensive experiments demonstrate that EMPFusion excels in infrared-visible fusion tasks within complex degraded scenes. Across the LLVIP, M3FD, RoadScene, and MSRS datasets, EMPFusion achieves state-of-the-art (SOTA) performance on multiple evaluation metrics, showcasing exceptional degradation robustness and visual-semantic information preservation capabilities. By unifying adaptive degradation correction with fusion, this research addresses fusion distortion caused by degraded multimodal data in harsh environments, significantly enhancing applicability and robustness in downstream tasks such as autonomous driving and security monitoring.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"298 ","pages":"Article 129844"},"PeriodicalIF":7.5,"publicationDate":"2025-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145221475","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 data-driven spatio-temporal driving risk field mechanism for path planning","authors":"Zhuoer Wang ,&nbsp;Baohan Shi ,&nbsp;Jianping Zhang ,&nbsp;Xiaowen Zhu ,&nbsp;Jian Zhou ,&nbsp;Bingrong Xu ,&nbsp;Bijun Li","doi":"10.1016/j.eswa.2025.129834","DOIUrl":"10.1016/j.eswa.2025.129834","url":null,"abstract":"<div><div>Characterizing the future risk posed by surrounding human-driven vehicles is crucial for enhancing the safety of autonomous vehicles. Existing risk field methods build spatiotemporal risk fields using mathematical models with fixed parameters, making them struggle to capture dynamic human driving behaviors such as frequent acceleration, deceleration or lane changes, and are prone to overlooking rare but critical sudden events, which leads to unstable risk assessments in complex long-term scenarios. To address the aforementioned issues, a data-driven spatio-temporal risk field framework is proposed, which builds on a Bidirectional Deep Ultra-Gated Recurrent Unit (BDUGRU) to capture the high-dimensional spatio-temporal features of nearby vehicles and precisely predict vehicle distribution patterns over extended horizons. The introduced approach manages to yield a more accurate risk field and significantly improves long-term risk assessment in complex traffic environments. Furthermore, to validate the model’s practicality in engineering, we integrated Rapidly-exploring Random Tree with spatiotemporal data-driven risk field (SRF-RRT) and conducted path-planning simulations for autonomous vehicles using real-world traffic data. The results demonstrate that the proposed model excels in both prediction accuracy and reliability, and effectively reduces the measurement error based on collision time (TTC), offering strong applicability and providing a novel theoretical foundation and technological route for path planning in intelligent connected vehicles (ICVs).</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"298 ","pages":"Article 129834"},"PeriodicalIF":7.5,"publicationDate":"2025-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145222091","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":"Training-free pedestrian trajectory prediction via segmentation-guided path planning","authors":"Dongchen Li,&nbsp;Zhimao Lin,&nbsp;Jinglu Hu","doi":"10.1016/j.eswa.2025.129770","DOIUrl":"10.1016/j.eswa.2025.129770","url":null,"abstract":"<div><div>Pedestrian trajectory prediction is a critical research topic for industrial applications and has been significantly advanced by deep learning. Despite decades of progress, current approaches still face two major challenges. First, the scarcity of data limits the generalization capability of deep learning models. Second, the absence of interpretability hinders real-world applications. To address these challenges, recent research has leveraged the knowledge of large language models (LLMs) to alleviate data sparsity and introduced novel knowledge-based methods to enhance interpretability. Nevertheless, transferring LLMs is excessively cumbersome, and the black-box nature of deep learning continues to obstruct interpretability. In our work, we propose a training-free transfer paradigm named Segmentation-Guided Path Planning (SGPP). Rather than directly transferring pretrained LLMs, SGPP introduces a more tailored and efficient transfer strategy by employing a promptable segmentation model. The segmentation model explicitly extracts walkable regions from the scenario, which serve as constraints in the planning space, thereby reformulating trajectory prediction as a more tractable white-box path-planning problem. Within this framework, our method offers a more effective solution to the two prevailing challenges. Compared with the latest training-free methods, our approach achieves superior performance and demonstrates strong generalization across diverse real-world scenarios, highlighting its suitability for industrial applications.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"298 ","pages":"Article 129770"},"PeriodicalIF":7.5,"publicationDate":"2025-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145222009","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":"Exploiting knowledge graph communities to fine-tune large language models","authors":"Alessia Amelio ,&nbsp;Christopher Buratti ,&nbsp;Michele Marchetti ,&nbsp;Davide Traini ,&nbsp;Domenico Ursino ,&nbsp;Luca Virgili","doi":"10.1016/j.eswa.2025.129816","DOIUrl":"10.1016/j.eswa.2025.129816","url":null,"abstract":"<div><div>Since the introduction of GPT-2, Large Language Models (LLMs) have proven to be able to handle various tasks with impressive performance. However, they sometimes generate incorrect output or even hallucinations. To overcome this problem, many researchers have investigated the possibility of integrating external factual knowledge, such as that encoded in Knowledge Graphs (KGs), into LLMs. Although there are many approaches in the existing literature that integrate KGs and LLMs in different ways, few of them use KGs to fine-tune LLMs, and none of them systematically use KG substructures. In this paper, we propose CoFine (Community-Based Fine-Tuner), an approach to fine-tune an LLM using the communities of a KG. CoFine works as follows: it first divides the KG into communities, each of which contains a homogeneous portion of the knowledge expressed by the KG. It then uses these communities to fine-tune the LLM. This way of proceeding allows LLM fine-tuning to focus on specific homogeneous information contained in the KG expressed by each community. CoFine allows the LLM to achieve a very high accuracy in knowledge completion tasks. This is evidenced by comparisons between CoFine and a baseline LLM fine-tuning approach, which showed that our approach achieves better results for all metrics considered with several KG.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"298 ","pages":"Article 129816"},"PeriodicalIF":7.5,"publicationDate":"2025-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145222007","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}