Expert Systems with Applications最新文献

{"title":"An offline-online collaborative optimization framework for the energy-efficient distributed hybrid flow shop scheduling problem with blocking constraints in electric anode carbon rod manufacturing system","authors":"Fuqing Zhao,&nbsp;Shangpeng Wang,&nbsp;Weiyuan Wang,&nbsp;Tianpeng Xu,&nbsp;NingNing Zhu","doi":"10.1016/j.eswa.2025.129955","DOIUrl":"10.1016/j.eswa.2025.129955","url":null,"abstract":"<div><div>The scheduling problem in the assembly workshop of carbon anodes for aluminum production is investigated within the distributed green manufacturing context. This scheduling problem is modeled as an Energy-Efficient Distributed Hybrid Flow Shop Scheduling Problem with Blocking Constraints (EEDHFSP-BC), with optimization objectives that include the minimization of both the makespan and the total energy consumption. To address this complex multi-objective optimization problem, an offline-online collaborative optimization framework (QLINSGA-II) integrating Q-Learning and an improved non-dominated sorting genetic algorithm (INSGA-II) is proposed. A two-phase offline-online scheduling strategy is adopted. First, a dedicated encoding scheme is designed according to the problem characteristics, and a hybrid initialization strategy is introduced during the offline learning phase. Meanwhile, three crossover and mutation operators integrating task assignment coordination and processing sequence allocation are developed to enhance global search capability. Second, a high-quality Pareto solution set is generated by INSGA-II, and Q-Learning is employed to learn from this solution set in the offline phase, thereby achieving intelligent guidance of the population evolution direction. Finally, trained agents are utilized in the online phase to dynamically adjust scheduling for newly arriving jobs. After the search process, a state evaluation mechanism is incorporated to dynamically guide the search by assessing the proportion of non-dominated solutions in the population, effectively improving the distribution and convergence of the Pareto solution set. Experimental results demonstrate that QLINSGA-II outperforms existing mainstream multi-objective optimization algorithms in terms of diversity, convergence speed, and solution coverage rate, providing an efficient and reliable solution for green workshop scheduling in the aluminum industry.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"298 ","pages":"Article 129955"},"PeriodicalIF":7.5,"publicationDate":"2025-10-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145268418","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":"Traditional Chinese medicine toxicity prediction by heterogeneous network","authors":"Yongzheng Zhu ,&nbsp;Yunbo Miao ,&nbsp;Rong Sun ,&nbsp;Zhongmin Yan ,&nbsp;Guoxian Yu","doi":"10.1016/j.eswa.2025.129969","DOIUrl":"10.1016/j.eswa.2025.129969","url":null,"abstract":"<div><div>The clinical usage of Traditional Chinese Medicines (TCMs) gains increasing international attention for their distinct therapeutic effects. Precisely understanding the herb (TCM) toxicity, and identifying toxic and effective herb ingredients are crucial for their safe use. Traditional wet-lab based pipelines for assessing herb toxicity are complex and time-consuming. Given the large volume toxicity data of ingredients, building computational models for efficient ingredient-based toxicity prediction and evaluation is promising, but often fails to effectively predict the toxicity of herbs, due to the complexity herbs and toxic mechanisms. To address these challenges, we propose a heterogeneous network based approach (HerbToxNet) for predicting herb toxicity. HerbToxNet first constructs a heterogeneous network composed with herbs, ingredients and molecular targets, and leverages a heterogeneous graph attention network to learn the representation of herbs. Meanwhile, it performs contrastive learning with dynamic coefficient to refine the representation by pulling close the herbs with shared toxic labels, while pushing away the others. Next, it uses Multilayer Perceptron (MLP) on the herb representation to predict the toxic labels of this herb, and further introduces a weighted label fusion strategy that uses toxic labels of similar herbs to augment the predicted labels of this herb. HerbToxNet outperforms competitive methods and finds out novel potential toxicities, with 96<span><math><mo>%</mo></math></span> toxicity labels confirmed for canonical herbs. It can mine related toxic ingredients and targets in an interpretable way, and dissect the molecular mechanism of herb toxicity with authenticity.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"299 ","pages":"Article 129969"},"PeriodicalIF":7.5,"publicationDate":"2025-10-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145334061","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 structure carved building mesh simplification","authors":"Zepeng Hou ,&nbsp;Fan Zhang ,&nbsp;Wenxuan Liu ,&nbsp;Yunlong Gao ,&nbsp;Xuan Wang ,&nbsp;Xianfeng Huang","doi":"10.1016/j.eswa.2025.129896","DOIUrl":"10.1016/j.eswa.2025.129896","url":null,"abstract":"<div><div>Three-dimensional building models are extensively utilized in domains such as smart cities and environmental assessments. However, the complex geometric structures and topological relationships of building models present significant challenges for applications such as rendering and spatial analysis. Existing methods rely on local geometry-based simplification approaches while underutilizing structural information, making it difficult to preserve both geometric structure fidelity and topological validity. This paper first proposes a novel approach to simplify man-made building meshes through structure carving. Initially, the spatial relationships of planar primitives are analyzed to construct an attribute-connected graph. The graph is then decomposed, and structures are extracted based on various connected relationships. Finally, the visual hull algorithm is employed to generate the visual mesh, which is subsequently simplified into a low-poly mesh through structure carving and mesh simplification. Structure carving introduces a novel framework comprising structure selection, structure primitives generation, and iterative carving, driven by depth loss. Experiments on various building models show that our method generates lightweight meshes that are watertight, accurate, and exhibit high geometric similarity, outperforming other state-of-the-art methods.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"299 ","pages":"Article 129896"},"PeriodicalIF":7.5,"publicationDate":"2025-10-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145334057","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":"Rootex 2.0: Multi-head deep learning and graph-based analysis for automated barley root phenotyping","authors":"Maichol Dadi,&nbsp;Annalisa Franco,&nbsp;Alessandra Lumini","doi":"10.1016/j.eswa.2025.129930","DOIUrl":"10.1016/j.eswa.2025.129930","url":null,"abstract":"<div><div>Understanding plant root architecture under diverse environmental conditions is crucial for improving crop resilience and ensuring global food security. We present a fully automated method for segmenting barley root systems from high-resolution images and detecting keypoints such as tips and sources with high precision. At the core of our approach is <em>DeepRoot-3H</em>, a novel multi-head deep network built upon the DeepLabv3+ backbone, designed to jointly handle root segmentation and keypoint detection within a unified architecture. This integrated design enhances both the consistency and robustness of the outputs.</div><div>A dedicated post-processing stage further refines keypoint localization, effectively handling challenges such as dense root clusters and variability in image quality. The resulting predictions are then structured into a graph representation, on which a path-walking algorithm identifies biologically meaningful connections between tips and sources. This enables the generation of RSML files and the extraction of critical morphological traits.</div><div>To evaluate the system, we employ IoU and Dice scores for segmentation quality, alongside Euclidean and weighted distance metrics for tip and source detection. We also assess the biological consistency of the extracted traits—such as total root length, tortuosity, covered area, and outer angles—through correlation and discrepancy measures. Experimental results on a challenging benchmark dataset demonstrate significant improvements over existing techniques, confirming the effectiveness and reliability of our method for high-fidelity root system analysis.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"299 ","pages":"Article 129930"},"PeriodicalIF":7.5,"publicationDate":"2025-10-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145334108","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":"Dual-objective prediction of garment pressure and body contouring efficacy in shapewear: a hybrid GA-BP framework for sizing systems","authors":"Zhu Wenhui ,&nbsp;Wang Fenfen ,&nbsp;Wang Gehui ,&nbsp;Wang Yongrong","doi":"10.1016/j.eswa.2025.129946","DOIUrl":"10.1016/j.eswa.2025.129946","url":null,"abstract":"<div><div>Accurate shapewear sizing is essential to maximize wearer comfort and achieve effective body-contouring outcomes, yet remains challenging due to the complex, nonlinear relationship between anthropometric data, garment pressure distribution, and subjective comfort perception. This study presents a novel application of a hybrid intelligent model that integrates backpropagation (BP) neural networks with genetic algorithm (GA) optimization to enhance the precision of size recommendations. The GA-BP framework specifically tackles the limitations of conventional sizing methods by simultaneously optimizing initial weights, activation thresholds, and network topology, effectively escaping local minima and enhancing generalization for dual-objective prediction of garment pressure and contouring efficacy<em>.</em> Experimental results demonstrate superior predictive performance, achieving coefficients of determination (R²) of 0.9896 (training) and 0.9854 (testing), significantly outperforming standard BP networks (0.837). Crucially, the model predicts key ergonomic indicators (e.g., localized pressure values) correlated with comfort, validated through controlled user trials. To bridge research and practice, an intelligent decision support system (DSS) was developed, featuring a C#-based GUI and MATLAB backend. This system processes anthropometric data to provide personalized sizing recommendations through an intuitive user interface. The findings confirm the effectiveness of GA-BP optimization in garment sizing and propose a user-centric decision support framework that harmonizes garment pressure with ergonomic comfort in shapewear selection.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"299 ","pages":"Article 129946"},"PeriodicalIF":7.5,"publicationDate":"2025-10-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145334109","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":"Breast mass classification in 3D ABUS based on Laplace-Beltrami spectra and dual path CNN","authors":"Sepideh Barekatrezaei ,&nbsp;Ali Naderiparizi ,&nbsp;Ehsan Kozegar ,&nbsp;Javad Ghofrani ,&nbsp;Mohsen Soryani","doi":"10.1016/j.eswa.2025.129973","DOIUrl":"10.1016/j.eswa.2025.129973","url":null,"abstract":"<div><div>Breast cancer is the most common cancer among women and remains a leading cause of cancer-related mortality worldwide. Accurately classifying breast masses as benign or malignant is crucial for guiding treatment and reducing unnecessary interventions. In this paper, we propose a hybrid deep learning-based classification framework for automated three-dimensional breast ultrasound (3D ABUS) images. The system integrates three classification paths: Support Vector Machine (SVM), Extremely Randomized Trees (Extra Trees), and a novel deep neural network. The SVM and Extra Trees classifiers utilize handcrafted features, including radiomic descriptors and Laplace-Beltrami eigenvalues. In these models, to reduce dimensionality of the Laplace-Beltrami features and prevent overfitting, Isomap is employed for nonlinear dimensionality reduction. The proposed neural network processes a 3D patch around the mass and the corresponding mask through two parallel paths, utilizing convolutional and max-pooling layers. The extracted features from both branches are concatenated with the complete Laplace-Beltrami feature vector before being classified by fully connected layers. To combine the outputs of all three base models, we employ a histogram-based gradient-boosting stacking classifier. This <em>meta</em>-classifier learns nonlinear dependencies between classifiers and enhances the overall performance. Experimental evaluation was conducted on the public TDSC-ABUS dataset, comprising 200 annotated breast volumes. The training/validation set includes 75 malignant and 55 benign cases, while the test set contains 40 malignant and 30 benign cases. On the test set, the proposed system achieves 84.29% accuracy, 93.50% AUC, 97.50% sensitivity, and 87.64% F1-score. Compared to the best competing method, it improves accuracy by 8.58% and AUC by 4.58%.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"299 ","pages":"Article 129973"},"PeriodicalIF":7.5,"publicationDate":"2025-10-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145334170","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":"Predicting last-mile delivery route deviations using machine learning","authors":"Anna Konovalenko ,&nbsp;Lars Magnus Hvattum ,&nbsp;Kim Aleksander Hammer Iversen","doi":"10.1016/j.eswa.2025.129921","DOIUrl":"10.1016/j.eswa.2025.129921","url":null,"abstract":"<div><div>Route planning in last-mile delivery is a complex task with many challenges, directly impacting delivery efficiency and costs. Drivers often deviate from optimized planned routes based on their knowledge. Using the properties of machine learning, this study aims to determine whether machine learning techniques can effectively predict deviations by drivers from planned routes and quantify the extent of such deviations. We propose to predict route deviations by analyzing a logistics company’s historical data of planned and actual routes using deep neural networks, with the dataset made publicly available. Our methodology incorporates both regression and classification models. The regression model estimates the degree of deviation, while the classification model aims to predict whether the deviation from a planned route will exceed a given threshold, based on different deviation metrics. As the input, we leverage the sequential structure of the route with route properties and drivers information. The computational experiments explore extending the given input to the models and testing various state-of-art neural network architectures. Our results demonstrate strong performance on both tasks, with our models achieving <span><math><mrow><mn>9</mn><mo>−</mo><mn>19</mn><mo>%</mo></mrow></math></span> improvements in regression metrics and <span><math><mrow><mn>3</mn><mo>−</mo><mn>15</mn><mo>%</mo></mrow></math></span> improvements in classification metrics compared to specified benchmarks, with statistical tests confirming the significance of these improvements.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"298 ","pages":"Article 129921"},"PeriodicalIF":7.5,"publicationDate":"2025-10-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145267979","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":"iJTyper: An effective type inference framework for incomplete java codes by integrating constraint- and statistics-based methods","authors":"Zhixiang Chen ,&nbsp;Anji Li ,&nbsp;Neng Zhang ,&nbsp;Jianguo Chen ,&nbsp;Yuan Huang ,&nbsp;Zibin Zheng","doi":"10.1016/j.eswa.2025.129972","DOIUrl":"10.1016/j.eswa.2025.129972","url":null,"abstract":"<div><div>Inferring the types of APIs used in incomplete codes (also referred to as code snippets), e.g., those on Q&amp;A forums, is a prerequisite step required to work with the codes. Existing type inference methods proposed for incomplete Java codes can be primarily categorized as constraint-based or statistics-based. The former relies on a pre-built API knowledge base (KB) and the type constraints in code snippets, which imposes higher requirements on code syntax and thus suffers from low recall due to the syntactic limitation. The latter overcomes the syntactic limitation by learning statistical regularities from a code corpus, however it rarely employs the type constraints in code snippets, which may lead to low precision. In this paper, we propose an effective type inference framework, called iJTyper, for incomplete Java codes by integrating the complementary advantages of constraint- and statistics-based methods. For a code snippet, iJTyper first applies a constraint-based method and augments the code context with the inferred API types. Then, it applies a statistics-based method to the augmented code snippet. The types predicted for APIs are further used to improve the constraint-based method by reducing its pre-built KB. iJTyper iteratively executes both methods and performs the code context augmentation and KB reduction mechanisms until a termination condition is satisfied. The final inference results are produced by combining the results of both methods. We implemented a version of iJTyper by integrating two state-of-the-art methods, SnR and MLMTyper, and evaluated iJTyper on two open-source datasets. Results show that 1) iJTyper achieves the highest average precision/recall¹ of 97.3 % and 92.5 % on both datasets; 2) iJTyper improves the average recall of SnR and MLMTyper by at least 7.3 % and 27.4 %, respectively; and 3) iJTyper improves the average precision/recall of the recently popular language model, ChatGPT, by 3.2 % and 0.5 % on both datasets.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"299 ","pages":"Article 129972"},"PeriodicalIF":7.5,"publicationDate":"2025-10-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145334063","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 collaborative reasoning framework for large language models in long-context Q&A","authors":"Jiacheng Yao,&nbsp;Guoxiu He,&nbsp;Xin Xu","doi":"10.1016/j.eswa.2025.129960","DOIUrl":"10.1016/j.eswa.2025.129960","url":null,"abstract":"<div><div>Large Language Models (LLMs) often struggle with the <em>Lost in the Middle</em> phenomenon in long-context question answering (Q&amp;A). Existing solutions, such as modifying attention mechanisms or positional encodings, typically require retraining, which demands substantial computational resources. Other strategies, including long-term memory mechanisms and context processing, heavily rely on auxiliary components and fail to fundamentally enhance the LLM’s reasoning capabilities. To bridge this gap, this paper proposes a novel collaborative reasoning framework. Initially, the framework uses a retrieval-augmented generation (RAG) approach to generate a candidate answer from sentences relevant to the input question. Subsequently, a training-free Shadow-LLM is designed to supplement local sentence-level information from the long-context during the reasoning process to produce another candidate answer. Finally, a <em>one-out-of-two</em> selection strategy chooses the final answer based on the two candidates. Experiments on three long-context Q&amp;A datasets and three backbone LLMs show that our method raises the F1 score over the baselines by 2% to 18%. Notably, we find that activating only the 0th decoder layer of the LLM is sufficient for Shadow-LLM to operate at optimal performance, enabling efficient deployment without retraining. The code is available at <span><span>link</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"299 ","pages":"Article 129960"},"PeriodicalIF":7.5,"publicationDate":"2025-10-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145334175","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":"Diffinformer: Diffusion informer model for long sequence time-series forecasting","authors":"Jiacheng Li ,&nbsp;Wei Chen ,&nbsp;Yican Liu ,&nbsp;Junmei Yang ,&nbsp;Zhiheng Zhou ,&nbsp;Delu Zeng","doi":"10.1016/j.eswa.2025.129944","DOIUrl":"10.1016/j.eswa.2025.129944","url":null,"abstract":"<div><div>Long sequence time-series forecasting (LSTF) is a significant research area with wide-ranging applications in energy, transportation, meteorology, and finance. Current methods primarily rely on statistical machine learning and deep neural network techniques to model historical time series for long-term forecasting. In recent years, Transformer-based models have demonstrated outstanding performance in forecasting, but their high computational costs limit their application. The Informer model addresses issues of high computational complexity and the management of long sequence inputs and outputs. However, existing models still face prediction bottlenecks under limited computational resources. The powerful generative capability of diffusion models can significantly enhance time series forecasting. We propose the Diffinformer model, which utilizes generative models for forecasting. Specifically, it combines conditional diffusion models with the ProbSparse self-attention distilling mechanism of Informer and incorporates the output of the diffusion model into the decoder to capture distant dependencies of observations from the perspective of dynamic systems. Comprehensive experimental results across five large-scale datasets demonstrate that Diffinformer improves predictive accuracy and outperforms corresponding baselines, offering a novel solution to the LSTF problem.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"299 ","pages":"Article 129944"},"PeriodicalIF":7.5,"publicationDate":"2025-10-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145334698","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}