{"title":"A comprehensive study of UK climate change policies based on complex systems modeling: Evidence from 2001 to 2020","authors":"Weiyi Jiang , Yihang Hong , Chun Xia Yang","doi":"10.1016/j.eswa.2025.130073","DOIUrl":"10.1016/j.eswa.2025.130073","url":null,"abstract":"<div><div>The United Kingdom is a highly industrialized and economically developed country, with greenhouse gas (GHGs) emissions historically well above the global average, thereby significantly contributing to global climate change. Over the past two decades, the UK government has introduced a range of measures to mitigate emissions, such as enacting legislation and employing market-based mechanisms. Between 2001 and 2020, the number of climate policies issued by the government and think tanks increased by 73 and 171 times, respectively, accompanied by a 43% reduction in GHGs and a 44% reduction in CO<sub>2</sub> emissions. Although prior research has identified a negative correlation between policy quantity and emissions, the relative importance of specific policy themes remains unclear. In this study, we employed the Latent Dirichlet Allocation (LDA) model to identify key thematic topics within UK climate policies published from 2001 to 2020. Subsequently, Partial Least Squares (PLS) regression was applied to quantify the contribution of each policy theme to reductions in GHGs emissions. Our quantitative results show that policies emphasizing themes such as “climate”, “carbon”, “greenhouse”, “gas”, and “low” accounted for 73.3% of the total emission reduction. Furthermore, we modeled the policy system as a complex network to assess structural interactions. These findings provide actionable insights for policymakers seeking to design more effective climate strategies.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"299 ","pages":"Article 130073"},"PeriodicalIF":7.5,"publicationDate":"2025-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145364982","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}
Qi-chao Sun , Jun-qing Li , Xiao-long Chen , Zhong-zhi Yang , Ya-nan Wang , Zhao-sheng Du , Li Wei
{"title":"Solving electric vehicle routing problem with recharging and battery swapping using a collaborative decision attention network","authors":"Qi-chao Sun , Jun-qing Li , Xiao-long Chen , Zhong-zhi Yang , Ya-nan Wang , Zhao-sheng Du , Li Wei","doi":"10.1016/j.eswa.2025.130116","DOIUrl":"10.1016/j.eswa.2025.130116","url":null,"abstract":"<div><div>With the growing demand for electric vehicles (EVs) in logistics and transportation, long charging times and limited driving range have emerged as significant challenges. Battery swapping offers a faster alternative to conventional charging, reducing downtime but introducing additional costs. This study investigates the electric vehicle routing problem with recharging and battery swapping (EVRP-RBS), which requires balancing range constraints with cost-efficiency. To address this, we propose a collaborative decision attention network (CDAN) based on deep reinforcement learning (DRL). CDAN jointly optimizes routing and charging strategies by training an encoder-decoder structured policy network. The EVRP-RBS is formulated as a two-action Markov decision process. The encoder extracts features from customer nodes, charging stations, and the depot, embedding them separately into a high-dimensional space. A self-attention mechanism is employed to capture the internode relationships, producing a global representation for downstream decision-making tasks. To effectively coordinate route planning and energy replenishment, we introduce a dual-attention decoder, which integrates two specialized attention modules—one for routing decisions and another for charging or battery swapping decisions. This architecture enables efficient integration of routing and charging considerations, significantly enhancing solution quality. Extensive experiments demonstrate that CDAN achieves competitive performance compared to both traditional and DRL-based baselines while exhibiting strong generalizability. Notably, the charging decision module plays a critical role in improving the overall solution quality.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"299 ","pages":"Article 130116"},"PeriodicalIF":7.5,"publicationDate":"2025-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145364920","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":"Fraud detection based on GNNs with local augmentation and adaptive relation aggregation","authors":"Zhou Mengzhe , Chen Jindong , Zhang Wen , Yan Zhihua","doi":"10.1016/j.eswa.2025.130110","DOIUrl":"10.1016/j.eswa.2025.130110","url":null,"abstract":"<div><div>Fraud detection based on Graph Neural Networks (GNNs) relies on aggregating information from the local neighborhoods, but this mechanism is vulnerable to two adversarial tactics: feature camouflage where fraudsters manipulate node attributes to mimic benign users, and relation camouflage where they establish connections with benign entities to dilute suspicious signals. These camouflage strategies compromise GNNs’ discriminative capability by exploiting the neighborhood aggregation mechanism itself. To address this vulnerability, we propose a fraud detection method based on GNNs with Local Augmentation and Adaptive Relation Aggregation (GNN-LAARA). GNN-LAARA integrates three synergistic components: a conditional variational autoencoder (CVAE) that generates discriminative node representations to expose camouflaged patterns, a reinforcement learning-based neighbor selector that dynamically filters noisy connections, and a multi-relational attention aggregator that adaptively fuses heterogeneous relationships. The effectiveness of GNN-LAARA is validated by two real-world fraud detection datasets. Experimental evaluation on two real-world fraud detection datasets demonstrates that GNN-LAARA achieves significant performance improvements, with up to 2.24% enhancement in AUC over state-of-the-art methods. Ablation studies further confirm the individual contributions of each module to the overall detection capability.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"299 ","pages":"Article 130110"},"PeriodicalIF":7.5,"publicationDate":"2025-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145364842","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":"MGRNN for dynamic constrained quadratic programming with verification and applications","authors":"Songjie Huang, Guancheng Wang, Xiuchun Xiao","doi":"10.1016/j.eswa.2025.130034","DOIUrl":"10.1016/j.eswa.2025.130034","url":null,"abstract":"<div><div>Dynamic Constrained Quadratic Programming (DCQP) is at the core of problems such as portfolio optimization and robot control. However, for this dynamic problem, the Gradient Recurrent Neural Network (GRNN) suffers lag errors and the Zeroing Neural Network (ZNN) requires costly matrix inversion and derivative information. Therefore, this paper proposes a Modified Gradient Recurrent Neural Network (MGRNN) to address these limitations. Its core adaptive mechanism that retains the simplicity of explicit dynamic structure while eliminating dependencies on matrix inversion and derivative computation, thereby resolving lag errors. Moreover, theoretical analyses demonstrate that the MGRNN achieves finite-time convergence and exhibits robust performance. Besides, performance analysis validates that the MGRNN outperforms traditional GRNN by significantly reducing residuals in solving the DCQP problem. Moreover, noise tolerance experiments reveal that the MGRNN also delivers the smallest residuals and the fastest convergence among all compared models under bounded noise, confirming its superior robustness. Furthermore, its efficacy and practicality are verified through current computation in dynamic circuits with temperature-dependent resistors, as well as through applications to portfolio optimization and manipulator control. Consequently, these results collectively highlight the effectiveness and practicality of MGRNN in addressing dynamic optimization tasks, providing a robust and computationally lightweight solution for real-time applications.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"299 ","pages":"Article 130034"},"PeriodicalIF":7.5,"publicationDate":"2025-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145364981","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":"WDAE-GAN: A hybrid dual autoencoder and generative adversarial framework with wavelet denoising for credit card fraud detection","authors":"Masoud RezvaniNejad , Ali Sabzali Yameqani","doi":"10.1016/j.eswa.2025.130078","DOIUrl":"10.1016/j.eswa.2025.130078","url":null,"abstract":"<div><div>Credit card fraud detection remains a critical challenge due to the severe class imbalance in real-world transaction datasets, where fraudulent cases represent only a minute fraction of total records. This study proposes WDAE-GAN, a hybrid detection framework that combines a Wasserstein Generative Adversarial Network (WGAN), dual autoencoders, wavelet-based denoising, and CatBoost classification. In this approach, the WGAN generates realistic synthetic fraud samples to augment the minority class, while dual autoencoders learn distinct latent representations from normal and fraud-augmented data. The concatenated latent features are refined through wavelet denoising before final classification by CatBoost, enhancing feature quality and reducing noise. Experimental results on two benchmark datasets—the European Credit Card Fraud Detection dataset and the IEEE-CIS Fraud Detection dataset—demonstrate that WDAE-GAN achieves near-perfect detection performance. On the European dataset, the model achieved a recall of 0.9999, precision of 0.9999, F1-score of 0.9999, AUC of 1.0000, and AUPRC of 0.9994. On the IEEE-CIS dataset, WDAE-GAN obtained a recall of 0.9994, precision of 0.9994, F1-score of 0.9994, AUC of 1.0000, and AUPRC of 0.9999. These results show that the proposed model not only delivers exceptional detection accuracy but also performs competitively against state-of-the-art methods, effectively identifying rare fraud instances while maintaining an extremely low false-positive rate. This confirms WDAE-GAN’s robustness and scalability for real-world financial fraud detection applications.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"299 ","pages":"Article 130078"},"PeriodicalIF":7.5,"publicationDate":"2025-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145364916","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}
Hongyue Guo , Qianying Yang , Yating Yu , Lidong Wang , Peng Jia , Witold Pedrycz
{"title":"Prediction of iron ore inventory at ports: A decomposition-integration hybrid approach incorporating key influencing factors","authors":"Hongyue Guo , Qianying Yang , Yating Yu , Lidong Wang , Peng Jia , Witold Pedrycz","doi":"10.1016/j.eswa.2025.130041","DOIUrl":"10.1016/j.eswa.2025.130041","url":null,"abstract":"<div><div>An accurate prediction of iron ore inventory at ports is necessary for analyzing market trends, optimizing operational strategies, and avoiding supply chain risks. Considering that the iron ore inventory is complex and influenced by various factors, this study offers a novel decomposition-integration hybrid model to fully capture the underlying patterns in inventory data and improve prediction accuracy. First, three significant components are extracted from the raw inventory sequence to represent high-, mid-, and low-frequency features using CEEMDAN decomposition and sample entropy reconstruction. Then, after investigating the potential influencing factors and diverse characteristics of each frequency sequence, we individually develop the prediction models by incorporating different influencing factors. Finally, the individual models’ outputs are integrated to achieve the final prediction, fully capturing the impact of key influencing factors on the iron ore inventory data at ports. Empirical results based on the data from Qingdao Port illustrate that the established hybrid forecasting model yields ideal accuracy, with at least a 2.11% reduction in RMSE and a minimum 1.73% reduction in MAE compared with nine models, verifying its effectiveness in forecasting iron ore inventory at ports.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"299 ","pages":"Article 130041"},"PeriodicalIF":7.5,"publicationDate":"2025-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145364411","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}
Yi You , Jiachang Gu , Zida Chen , Gang Wu , Kang Gao
{"title":"MonoGuard: towards infrastructure-aware oversized vehicle detection via dynamic 3D metrology with monocular surveillance cameras","authors":"Yi You , Jiachang Gu , Zida Chen , Gang Wu , Kang Gao","doi":"10.1016/j.eswa.2025.130049","DOIUrl":"10.1016/j.eswa.2025.130049","url":null,"abstract":"<div><div>Oversized vehicles pose a significant threat to public safety and urban infrastructure, leading to catastrophic bridge failures and traffic disruptions worldwide. However, existing detection systems often lack the automation, real-time capability, and cost-effectiveness required for widespread deployment. To address this gap, this study introduces MonoGuard, a deep learning-enhanced framework for real-time oversized vehicle detection (OSVD). By leveraging ubiquitous monocular surveillance cameras, Monoguard is designed to guard against infrastructure damage and enhance traffic safety. Monoguard integrates three key innovations: a semi-automatic calibration workflow using Segment Anything Model 2 (SAM2) and multi-frame fusion for robust vanishing point estimation; the lightweight Context-Guided Attention Segmentation-You Only Look Once (CGAS-YOLO) model for efficient vehicle segmentation; and a rule-based adaptive 3D bounding box pipeline that dynamically adjusts to camera-object geometry. Extensive evaluations across 36 UAV-simulated scenarios demonstrate MonoGuard’s high performance. It achieves a remarkable height estimation accuracy rate of 96.98% for cars and 95.50% for trucks, while maintaining a real-time throughput of 46.5 FPS on an RTX 4080 Laptop. By repurposing existing surveillance infrastructure, MonoGuard provides a scalable and economical solution for smart cities, enabling early warnings to prevent collisions, protect infrastructure, and safeguard lives and property.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"299 ","pages":"Article 130049"},"PeriodicalIF":7.5,"publicationDate":"2025-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145334034","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}
Juan Zhou , Qianwang Deng , Yinwen Ma , Rui Pan , Jingxing Zhang , Mao Tan
{"title":"Double-layer Q-learning guided NSGA-II for integrated production scheduling and inventory decision considering multi-product orders","authors":"Juan Zhou , Qianwang Deng , Yinwen Ma , Rui Pan , Jingxing Zhang , Mao Tan","doi":"10.1016/j.eswa.2025.130008","DOIUrl":"10.1016/j.eswa.2025.130008","url":null,"abstract":"<div><div>High-quality spare parts production and supply are critical for establishing sustained competitive advantages. However, limited manufacturing resources and tight delivery deadlines challenge enterprises in balancing delivery reliability and inventory capital occupation. The widespread existence of multi-product spare parts orders, characterized by interdependent delivery constraints, further complicates production resource allocation and operational coordination. To address these issues, this study treats production and inventory as dual sources for order fulfillment and investigates the integrated production scheduling and inventory decision optimization problem considering multi-product orders (IPSID-MPO), aiming to simultaneously minimize total inventory capital occupation and total order delay penalty. We propose a novel double-layer Q-learning-guided non-dominated sorting genetic algorithm-II (DQ-NSGA), incorporating key innovations: (i) a decoding strategy that considers product delivery constraints, (ii) a hybrid initialization mechanism integrating four problem-specific heuristics, (iii) knowledge-driven local search strategies, and (iv) a self-adaptive adjustment mechanism via double-layer Q-learning. The outer layer dynamically tunes crossover/mutation probabilities based on population evolution, while the inner layer guides individual-specific search strategies. Comprehensive experiments on 180 benchmark instances demonstrate DQ-NSGA’s superiority over mainstream comparative algorithms. Comparisons with common simplified models that consider single-product orders demonstrates the necessity of incorporating multi-product orders in production and inventory decision-making. Furthermore, compared to Make-to-Stock and Make-to-Order paradigms, the proposed IPSID-MPO model not only reduces inventory carrying costs by 28.7% but also effectively enhances the flexibility of the manufacturing system.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"299 ","pages":"Article 130008"},"PeriodicalIF":7.5,"publicationDate":"2025-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145334029","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}
Mohamed Assaf , Sena Assaf , Xinming Li , Mohamed Al-Hussein
{"title":"A multi-user game-based system for planning modular construction activities","authors":"Mohamed Assaf , Sena Assaf , Xinming Li , Mohamed Al-Hussein","doi":"10.1016/j.eswa.2025.130050","DOIUrl":"10.1016/j.eswa.2025.130050","url":null,"abstract":"<div><div>Supply chain (SC) planning in modular construction (MC) can be challenging because it requires interconnected and complex activities among various teams and across different project stages. Recently, game engines have been increasingly used to resolve these challenges, as they create realistic virtual environments and simulations of possible scenarios before actual project implementation. However, game engine applications have been restricted to single-user and centralized models, limiting real-time collaboration among MC teams. Within a Design Science Research methodology, this study proposes an intelligent game-based modular planning (GAMMOD) system, supported by multi-user functions, which is flexible in terms of access, allowing for either non-immersive or immersive mode, depending on the available hardware tools, for collaborative planning of the MC-SC. The GAMMOD system integrates a blockchain protocol for data security in the non-immersive mode, while the immersive mode relies on user credentials authorization. The GAMMOD system considers both numerical key performance indicators, such as sustainability, cost, and time, as well as practical ones, including road dimensions, module clearance, and possible clashes. Two distinct case studies, representing different MC types, are presented to illustrate the features of the GAMMOD system. The evaluation tests of the GAMMOD system conducted by 14 MC experts have shown a general consensus on its functionality, with 80% to 100% of the participants agreeing or strongly agreeing on the GAMMOD system’s performance. Additionally, the GAMMOD system demonstrated a usability score of 75.7, surpassing the established threshold of 70. The GAMMOD system is expected to help MC stakeholders make informed, collaborative decisions and develop a shared understanding of decision feasibility, potential conflicts, and constraints before the commencement of the MC project.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"299 ","pages":"Article 130050"},"PeriodicalIF":7.5,"publicationDate":"2025-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145364413","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}
Xiaoyan Su , Xuying Huang , Xiaolei Pan , Debiao Meng
{"title":"A dependence assessment method based on quantum model of mass function in human reliability analysis","authors":"Xiaoyan Su , Xuying Huang , Xiaolei Pan , Debiao Meng","doi":"10.1016/j.eswa.2025.129992","DOIUrl":"10.1016/j.eswa.2025.129992","url":null,"abstract":"<div><div>Human reliability analysis (HRA) has garnered widespread attention in high-reliability-demand fields such as the nuclear industry. The dependence assessment among human failure events (HFEs) constitutes a crucial component of HRA research, as it enhances the accuracy of HRA outcomes and contributes to reducing human error probabilities. This paper proposes a novel method based on the quantum model of mass function (QMMF) to address dependence assessment in HRA under uncertain dynamic scenarios. Firstly, dependence influencing factors are identified and their basic belief assignments (BBAs) are constructed based on expert evaluations. Then, a time correction model is developed to generate time-corrected BBAs, upon which the QMMF is applied to reconstruct dynamic factor BBAs. Finally, the conditional human error probability (CHEP) is calculated through the fusion of reconstructed dynamic factor BBAs and static factor BBAs. The proposed method, grounded in quantum evidence theory, assigns the physical meaning of “time” to the “phase angle” variable, enabling flexible expression of evidence evolution over time while maintaining solid theoretical foundations and compatibility. Additionally, the method allows adjustment of temporal correction intensity for dynamic factors by modifying the weight distribution in time-corrected BBAs. Case study results demonstrate that the proposed method can yield more accurate and rational outcomes.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"299 ","pages":"Article 129992"},"PeriodicalIF":7.5,"publicationDate":"2025-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145334110","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}