Expert Systems with Applications最新文献

MSIFT: A novel end-to-end mechanical fault diagnosis framework under limited & imbalanced data using multi-source information fusion

IF 7.5 1区计算机科学 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Expert Systems with Applications

Pub Date : 2025-02-24 DOI: 10.1016/j.eswa.2025.126947

Yue Yu , Hamid Reza Karimi , Len Gelman , Ahmet Enis Cetin

Data-driven intelligent fault diagnosis methods have emerged as powerful tools for monitoring and maintaining the operating conditions of mechanical equipment. However, in real-world engineering scenarios, mechanical equipment typically operates under normal conditions, resulting in limited and imbalanced (L&I) data. This situation gives rise to label bias and biased training. Meanwhile, the current multi-source information fault diagnosis research to date has tended to focus on fault identification rather than effective feature fusion strategies. To solve these issues, a novel end-to-end mechanical fault diagnosis framework under limited & imbalanced data using multi-source information fusion is proposed to model data-level and algorithm-level ideas in a unified deep network for achieving effective multi-source information fusion under the L&I working conditions. From a data-level perspective, a data preprocessing operation is first employed to capture time–frequency information simultaneously. Subsequently, multi-source time–frequency information is fed into feature extractors with information discriminators to construct local and information-invariant feature maps with different scales to eliminate multi-source information domain shift. Then, the multi-source feature vectors are modeled by a multi-source information transformer-based neural network to achieve effective multi-source information fusion through cross-attention mechanism. Next, the global max pooling and global average pooling layers are leveraged to obtain the more representative features. Finally, from an algorithm-level perspective, a dual-stream diagnosis predictor with a binary diagnosis predictor and a multi-class diagnosis predictor is designed to synthesize the diagnostic results through a reweighing activation mechanism for addressing the L&I problems. Extensive experiments on four different multi-source information datasets show the superiority and promising performance of our method compared to the state-of-the-art methods, as evidenced by indicators from various aspects.

{"title":"MSIFT: A novel end-to-end mechanical fault diagnosis framework under limited & imbalanced data using multi-source information fusion","authors":"Yue Yu , Hamid Reza Karimi , Len Gelman , Ahmet Enis Cetin","doi":"10.1016/j.eswa.2025.126947","DOIUrl":"10.1016/j.eswa.2025.126947","url":null,"abstract":"<div><div>Data-driven intelligent fault diagnosis methods have emerged as powerful tools for monitoring and maintaining the operating conditions of mechanical equipment. However, in real-world engineering scenarios, mechanical equipment typically operates under normal conditions, resulting in limited and imbalanced (L&I) data. This situation gives rise to label bias and biased training. Meanwhile, the current multi-source information fault diagnosis research to date has tended to focus on fault identification rather than effective feature fusion strategies. To solve these issues, a novel end-to-end mechanical fault diagnosis framework under limited & imbalanced data using multi-source information fusion is proposed to model data-level and algorithm-level ideas in a unified deep network for achieving effective multi-source information fusion under the L&I working conditions. From a data-level perspective, a data preprocessing operation is first employed to capture time–frequency information simultaneously. Subsequently, multi-source time–frequency information is fed into feature extractors with information discriminators to construct local and information-invariant feature maps with different scales to eliminate multi-source information domain shift. Then, the multi-source feature vectors are modeled by a multi-source information transformer-based neural network to achieve effective multi-source information fusion through cross-attention mechanism. Next, the global max pooling and global average pooling layers are leveraged to obtain the more representative features. Finally, from an algorithm-level perspective, a dual-stream diagnosis predictor with a binary diagnosis predictor and a multi-class diagnosis predictor is designed to synthesize the diagnostic results through a reweighing activation mechanism for addressing the L&I problems. Extensive experiments on four different multi-source information datasets show the superiority and promising performance of our method compared to the state-of-the-art methods, as evidenced by indicators from various aspects.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"274 ","pages":"Article 126947"},"PeriodicalIF":7.5,"publicationDate":"2025-02-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143480048","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Developing a decision support system using different classification algorithms for polyclinic selection

IF 7.5 1区计算机科学 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Expert Systems with Applications

Pub Date : 2025-02-24 DOI: 10.1016/j.eswa.2025.127042

Müberra Terzi Kumandaş , Naci Murat

A significant part of the patients applying to the emergency department in Turkey are green triage patients. Green triage means patients keep the emergency department unnecessarily busy. This situation causes inefficient use of health services and unnecessary density in the emergency department. This study aims to create a decision support system that allows patients to be directed to the right polyclinic using text-mining techniques and a large language model (LLM). The study sample consists of medical records of patients who came to the emergency department within a year. The study was carried out in two steps: association analysis and classification analysis. Zemberek Natural Language Library was used for root analysis of the words in the data set. 32 association rules were obtained from the data with the Apriori algorithm. Classification analysis was performed for word-polyclinic matching according to association analysis rules. Of the classification algorithms used decision tree, k-nearest neighbors (K-NN), support vector machines (SVM), and random forest. Accuracy rates were obtained as 81.3 %, 79.6 %, 83.4 % and 83.1 %, respectively. Additionally, the classification was performed using ChatGPT from LLMs. Polyclinic classification made with ChatGPT was found 78.9 % accuracy rate. All classical machine learning algorithms showed higher accuracy than ChatGPT. However, when ChatGPT’s Cohen’s kappa (0.798) and F-measure (0.813) values are examined, it can be said that it is similar to the Random Forest algorithm and the SVM algorithm. Nevertheless, the highest accuracy rate belongs to the SVM algorithm. This study shows that the SVM algorithm can classify patients on a polyclinic basis according to their complaints and that an effective decision support system that helps guide patients can be created.

在土耳其急诊科就诊的病人中，有很大一部分是绿色分流病人。绿色分流意味着患者让急诊科处于不必要的忙碌状态。这种情况会导致医疗服务的低效利用和急诊科不必要的密度。本研究旨在创建一个决策支持系统，利用文本挖掘技术和大型语言模型（LLM）将患者引导到合适的综合医院。研究样本包括一年内急诊科就诊患者的医疗记录。研究分两步进行：关联分析和分类分析。Zemberek 自然语言库用于对数据集中的词根进行分析。利用 Apriori 算法从数据中获得了 32 条关联规则。分类分析是根据关联分析规则进行词-多环匹配。分类算法包括决策树、k-近邻（K-NN）、支持向量机（SVM）和随机森林。准确率分别为 81.3%、79.6%、83.4% 和 83.1%。此外，还使用来自 LLMs 的 ChatGPT 进行了分类。使用 ChatGPT 进行的综合门诊分类准确率为 78.9%。所有经典机器学习算法的准确率都高于 ChatGPT。不过，如果对 ChatGPT 的 Cohen's kappa (0.798) 和 F-measure (0.813) 值进行检验，可以说它与随机森林算法和 SVM 算法类似。不过，SVM 算法的准确率最高。这项研究表明，SVM 算法可以根据患者的主诉对综合门诊的患者进行分类，并可以创建一个有效的决策支持系统来帮助指导患者。

{"title":"Developing a decision support system using different classification algorithms for polyclinic selection","authors":"Müberra Terzi Kumandaş , Naci Murat","doi":"10.1016/j.eswa.2025.127042","DOIUrl":"10.1016/j.eswa.2025.127042","url":null,"abstract":"<div><div>A significant part of the patients applying to the emergency department in Turkey are green triage patients. Green triage means patients keep the emergency department unnecessarily busy. This situation causes inefficient use of health services and unnecessary density in the emergency department. This study aims to create a decision support system that allows patients to be directed to the right polyclinic using text-mining techniques and a large language model (LLM). The study sample consists of medical records of patients who came to the emergency department within a year. The study was carried out in two steps: association analysis and classification analysis. Zemberek Natural Language Library was used for root analysis of the words in the data set. 32 association rules were obtained from the data with the Apriori algorithm. Classification analysis was performed for word-polyclinic matching according to association analysis rules. Of the classification algorithms used decision tree, k-nearest neighbors (K-NN), support vector machines (SVM), and random forest. Accuracy rates were obtained as 81.3 %, 79.6 %, 83.4 % and 83.1 %, respectively. Additionally, the classification was performed using ChatGPT from LLMs. Polyclinic classification made with ChatGPT was found 78.9 % accuracy rate. All classical machine learning algorithms showed higher accuracy than ChatGPT. However, when ChatGPT’s Cohen’s kappa (0.798) and F-measure (0.813) values are examined, it can be said that it is similar to the Random Forest algorithm and the SVM algorithm. Nevertheless, the highest accuracy rate belongs to the SVM algorithm. This study shows that the SVM algorithm can classify patients on a polyclinic basis according to their complaints and that an effective decision support system that helps guide patients can be created.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"274 ","pages":"Article 127042"},"PeriodicalIF":7.5,"publicationDate":"2025-02-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143488656","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}

引用次数: 0

Exploring multi-scale and cross-type features in 3D point cloud learning with CCMNET

IF 7.5 1区计算机科学 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Expert Systems with Applications

Pub Date : 2025-02-24 DOI: 10.1016/j.eswa.2025.126960

Wei Zhou , Weiwei Jin , Dekui Wang , Xingxing Hao , Yongxiang Yu , Caiwen Ma

The existing methods for 3D point cloud learning can be broadly categorized into point-based and voxel-based approaches. Typically, these techniques often produce features that are either overly fine-grained or excessively coarse-grained. Moreover, most conventional methods primarily concentrate on extracting multi-scale information from a single feature type, overlooking the potential advantages of integrating diverse multi-scale features. To overcome these limitations, we propose CCMNet, an innovative framework for 3D point cloud learning that leverages Coarse-to-fine and Cross-type Multi-scale features. CCMNet integrates three levels of feature granularity: coarse-grained, mid-grained, and fine-grained. Coarse-grained features are extracted using a 3D CNN with low voxel resolution, mid-grained features are captured through an attention mechanism operating both within and across neighborhoods, and fine-grained features are derived using a streamlined multi-layer perceptron (MLP) network. In addition, we introduce a cross-type multi-scale strategy to enhance local feature representations by seamlessly integrating features across different scales and types. CCMNet serves as the feature extraction network for point cloud classification and segmentation tasks. Experimental results highlight that our method achieves significant performance improvements in 3D point cloud learning. The source code is publicly available at https://github.com/NWUzhouwei/CCMNet.

{"title":"Exploring multi-scale and cross-type features in 3D point cloud learning with CCMNET","authors":"Wei Zhou , Weiwei Jin , Dekui Wang , Xingxing Hao , Yongxiang Yu , Caiwen Ma","doi":"10.1016/j.eswa.2025.126960","DOIUrl":"10.1016/j.eswa.2025.126960","url":null,"abstract":"<div><div>The existing methods for 3D point cloud learning can be broadly categorized into point-based and voxel-based approaches. Typically, these techniques often produce features that are either overly fine-grained or excessively coarse-grained. Moreover, most conventional methods primarily concentrate on extracting multi-scale information from a single feature type, overlooking the potential advantages of integrating diverse multi-scale features. To overcome these limitations, we propose <strong>CCMNet</strong>, an innovative framework for 3D point cloud learning that leverages <u><strong>C</strong></u>oarse-to-fine and <u><strong>C</strong></u>ross-type <u><strong>M</strong></u>ulti-scale features. CCMNet integrates three levels of feature granularity: coarse-grained, mid-grained, and fine-grained. Coarse-grained features are extracted using a 3D CNN with low voxel resolution, mid-grained features are captured through an attention mechanism operating both within and across neighborhoods, and fine-grained features are derived using a streamlined multi-layer perceptron (MLP) network. In addition, we introduce a cross-type multi-scale strategy to enhance local feature representations by seamlessly integrating features across different scales and types. CCMNet serves as the feature extraction network for point cloud classification and segmentation tasks. Experimental results highlight that our method achieves significant performance improvements in 3D point cloud learning. The source code is publicly available at <span><span>https://github.com/NWUzhouwei/CCMNet</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"274 ","pages":"Article 126960"},"PeriodicalIF":7.5,"publicationDate":"2025-02-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143488615","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}

引用次数: 0

Advanced deep learning model for crop-specific and cross-crop pest identification

IF 7.5 1区计算机科学 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Expert Systems with Applications

Pub Date : 2025-02-24 DOI: 10.1016/j.eswa.2025.126896

Md Suzauddola , Defu Zhang , Adnan Zeb , Junde Chen , Linsen Wei , A.B.M. Sadique Rayhan

The agricultural pests across diverse crops pose significant challenges in identification due to their diminutive size, natural camouflage, and the complex, cluttered environments they inhabit. This paper proposes an advanced deep-learning model to address these issues. The Key components of this customized solution include the “Channel-Enhanced Generalized Efficient Layer Aggregation Network” module, which enhances and highlights the features through the channel and spatial enhancement mechanisms. The “Generalized Multi-Scale Feature Extraction” module employs multi-scale feature extraction to provide fine-grain, inter-scale, and rich pest features. Additionally, a custom re-parameterization technique was adapted to optimize the real-time performance and boost the model’s efficiency. The model’s effectiveness was rigorously evaluated using the proposed Jute17 dataset. Experimental results demonstrate significant performance improvements over the Baseline model, achieving a 9.2% increase in Precision, and the detection speed retains high efficiency on the Jute17 dataset. Furthermore, the benchmark datasets Pest24 and IP102 were added to validate the performance of the proposed model, and it outperformed the Baseline model, Faster RCNN, Deformable-detr, and other YOLO series. The proposed model attained a mean Average Precision (mAP) of 78.22% on the Pest24 dataset and 78.15% accuracy on the IP102 dataset. This method offers a practical and efficient agricultural crop-specific and cross-crop pest management solution for complex field environments.

{"title":"Advanced deep learning model for crop-specific and cross-crop pest identification","authors":"Md Suzauddola , Defu Zhang , Adnan Zeb , Junde Chen , Linsen Wei , A.B.M. Sadique Rayhan","doi":"10.1016/j.eswa.2025.126896","DOIUrl":"10.1016/j.eswa.2025.126896","url":null,"abstract":"<div><div>The agricultural pests across diverse crops pose significant challenges in identification due to their diminutive size, natural camouflage, and the complex, cluttered environments they inhabit. This paper proposes an advanced deep-learning model to address these issues. The Key components of this customized solution include the “Channel-Enhanced Generalized Efficient Layer Aggregation Network” module, which enhances and highlights the features through the channel and spatial enhancement mechanisms. The “Generalized Multi-Scale Feature Extraction” module employs multi-scale feature extraction to provide fine-grain, inter-scale, and rich pest features. Additionally, a custom re-parameterization technique was adapted to optimize the real-time performance and boost the model’s efficiency. The model’s effectiveness was rigorously evaluated using the proposed Jute17 dataset. Experimental results demonstrate significant performance improvements over the Baseline model, achieving a 9.2% increase in Precision, and the detection speed retains high efficiency on the Jute17 dataset. Furthermore, the benchmark datasets Pest24 and IP102 were added to validate the performance of the proposed model, and it outperformed the Baseline model, Faster RCNN, Deformable-detr, and other YOLO series. The proposed model attained a mean Average Precision (<em>mAP</em>) of 78.22% on the Pest24 dataset and 78.15% accuracy on the IP102 dataset. This method offers a practical and efficient agricultural crop-specific and cross-crop pest management solution for complex field environments.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"274 ","pages":"Article 126896"},"PeriodicalIF":7.5,"publicationDate":"2025-02-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143480047","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}

引用次数: 0

An enhanced combined model for water quality prediction utilizing spatiotemporal features and physical-informed constraints

IF 7.5 1区计算机科学 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Expert Systems with Applications

Pub Date : 2025-02-24 DOI: 10.1016/j.eswa.2025.126937

Jiaming Zhu , Wan Dai , Jingyi Shao , Jinpei Liu , Huayou Chen

Water quality is vital for both human health and the ecological environment, and accurate predictions of the Water Quality Index (WQI) play a key role in timely monitoring of water conditions, providing essential support for environmental protection and management efforts. However, most existing studies focus on single monitoring sites, overlooking the interactions between neighboring locations. To address this limitation, this paper proposes a comprehensive water quality prediction framework that integrates spatial feature extraction, ordinary differential equations, and physical-informed constraints, called PI-GCN-SRLSTM. This framework enables WQI predictions across multiple sites and future time steps. First, a two-layer Graph Convolutional Network (GCN) is employed to capture the spatial features of WQI. Next, an enhanced Long Short-Term Memory network (LSTM), based on a second-order residual network (SRLSTM), is designed to capture complex temporal dynamics. Finally, to ensure predictions remain realistic, a gradient constraint is incorporated into the model’s loss function, improving the stability and reliability of the results. Experimental results based on the Chaohu lake dataset demonstrate that the proposed framework outperforms state-of-the-art benchmark models across six evaluation metrics: RMSE, MSE, MAE, MAPE, m1, and m2, with improvements of 30.47%, 43.04%, 29.57%, 29.92%, 0.83%, and 1.63%, respectively.

{"title":"An enhanced combined model for water quality prediction utilizing spatiotemporal features and physical-informed constraints","authors":"Jiaming Zhu , Wan Dai , Jingyi Shao , Jinpei Liu , Huayou Chen","doi":"10.1016/j.eswa.2025.126937","DOIUrl":"10.1016/j.eswa.2025.126937","url":null,"abstract":"<div><div>Water quality is vital for both human health and the ecological environment, and accurate predictions of the Water Quality Index (WQI) play a key role in timely monitoring of water conditions, providing essential support for environmental protection and management efforts. However, most existing studies focus on single monitoring sites, overlooking the interactions between neighboring locations. To address this limitation, this paper proposes a comprehensive water quality prediction framework that integrates spatial feature extraction, ordinary differential equations, and physical-informed constraints, called PI-GCN-SRLSTM. This framework enables WQI predictions across multiple sites and future time steps. First, a two-layer Graph Convolutional Network (GCN) is employed to capture the spatial features of WQI. Next, an enhanced Long Short-Term Memory network (LSTM), based on a second-order residual network (SRLSTM), is designed to capture complex temporal dynamics. Finally, to ensure predictions remain realistic, a gradient constraint is incorporated into the model’s loss function, improving the stability and reliability of the results. Experimental results based on the Chaohu lake dataset demonstrate that the proposed framework outperforms state-of-the-art benchmark models across six evaluation metrics: RMSE, MSE, MAE, MAPE, m1, and m2, with improvements of 30.47%, 43.04%, 29.57%, 29.92%, 0.83%, and 1.63%, respectively.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"274 ","pages":"Article 126937"},"PeriodicalIF":7.5,"publicationDate":"2025-02-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143488657","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}

引用次数: 0

Research on improving the robustness of spatially embedded interdependent networks by adding local additional dependency links

IF 7.5 1区计算机科学 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Expert Systems with Applications

Pub Date : 2025-02-24 DOI: 10.1016/j.eswa.2025.127035

Jiaqi Liang , Zhengcheng Dong , Meng Tian , June Li

Most of the existing research on interdependent networks always focuses on topology, without considering the component spatial information. However, for some real interdependent infrastructures, the dependency links are more likely established locally rather than globally. In this paper, we investigate the effects of local coupling patterns impacting on the robustness of interdependent networks considering spatial information. Firstly, the interdependent scale-free network is located in a 2-D square unit plane where dependent nodes falling into a circle

O_{c}

with radius r are connected. Secondly, three novel local coupling patterns, including local low degree–degree coupling, local neighbor node coupling and local random coupling, are introduced. To verify they have better effects on improving the robustness, the traditional global low degree–degree coupling, global neighbor node coupling and global random coupling are selected as comparing patterns. Finally, the improving effect rank order of proposed local coupling patterns and the equilibrium point between r and effects are obtained. Specifically, under topological attacks, with the increase of r, the effects of LLD, LNN and LR always have better performance, and the equilibrium point is

r = 0.2

, when

0.2 < r \leq 1.4

, the effects cannot be improved obviously. Under localized attacks, besides local coupling patterns are better than global ones, the equilibrium point is

r = 0.3

, when

0.3 < r \leq 0.8

, the effects are improved faintly. With r from 0.8 to 1.4, the effects remain constants. These findings can be as a reference to improving some real interdependent infrastructures.

{"title":"Research on improving the robustness of spatially embedded interdependent networks by adding local additional dependency links","authors":"Jiaqi Liang , Zhengcheng Dong , Meng Tian , June Li","doi":"10.1016/j.eswa.2025.127035","DOIUrl":"10.1016/j.eswa.2025.127035","url":null,"abstract":"<div><div>Most of the existing research on interdependent networks always focuses on topology, without considering the component spatial information. However, for some real interdependent infrastructures, the dependency links are more likely established locally rather than globally. In this paper, we investigate the effects of local coupling patterns impacting on the robustness of interdependent networks considering spatial information. Firstly, the interdependent scale-free network is located in a 2-D square unit plane where dependent nodes falling into a circle <span><math><msub><mi>O</mi><mi>c</mi></msub></math></span> with radius <em>r</em> are connected. Secondly, three novel local coupling patterns, including local low degree–degree coupling, local neighbor node coupling and local random coupling, are introduced. To verify they have better effects on improving the robustness, the traditional global low degree–degree coupling, global neighbor node coupling and global random coupling are selected as comparing patterns. Finally, the improving effect rank order of proposed local coupling patterns and the equilibrium point between <em>r</em> and effects are obtained. Specifically, under topological attacks, with the increase of <em>r</em>, the effects of LLD, LNN and LR always have better performance, and the equilibrium point is <span><math><mrow><mi>r</mi><mo>=</mo><mn>0.2</mn></mrow></math></span>, when <span><math><mrow><mn>0.2</mn><mo><</mo><mi>r</mi><mo>≤</mo><mn>1.4</mn></mrow></math></span>, the effects cannot be improved obviously. Under localized attacks, besides local coupling patterns are better than global ones, the equilibrium point is <span><math><mrow><mi>r</mi><mo>=</mo><mn>0.3</mn></mrow></math></span>, when <span><math><mrow><mn>0.3</mn><mo><</mo><mi>r</mi><mo>≤</mo><mn>0.8</mn></mrow></math></span>, the effects are improved faintly. With <em>r</em> from 0.8 to 1.4, the effects remain constants. These findings can be as a reference to improving some real interdependent infrastructures.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"274 ","pages":"Article 127035"},"PeriodicalIF":7.5,"publicationDate":"2025-02-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143488617","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}

引用次数: 0

Referring flexible image restoration

IF 7.5 1区计算机科学 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Expert Systems with Applications

Pub Date : 2025-02-24 DOI: 10.1016/j.eswa.2025.126857

Runwei Guan , Rongsheng Hu , Zhuhao Zhou , Tianlang Xue , Ka Lok Man , Jeremy Smith , Eng Gee Lim , Weiping Ding , Yutao Yue

In reality, images often exhibit multiple degradations, such as rain and fog at night (triple degradations). However, in many cases, individuals may not want to remove all degradations, for instance, a blurry lens revealing a beautiful snowy landscape (double degradations). In such scenarios, people may only desire to deblur. These situations and requirements shed light on a new challenge in image restoration, where a model must perceive and remove specific degradation types specified by human commands in images with multiple degradations. We term this task Referring Flexible Image Restoration (RFIR). To address this, we first construct a large-scale synthetic dataset called RFIR, comprising 153,423 samples with the degraded image, text prompt for specific degradation removal and restored image. RFIR consists of five basic degradation types: blur, rain, haze, low light and snow while six main sub-categories are included for varying degrees of degradation removal. To tackle the challenge, we propose a novel transformer-based multi-task model named TransRFIR, which simultaneously perceives degradation types in the degraded image and removes specific degradation upon text prompt. TransRFIR is based on two devised modules, Multi-Head Agent Self-Attention (MHASA) for multi-degradation context modeling and Multi-Head Agent Cross Attention (MHACA) for efficient alignment between prompt and referred degradations, where MHASA and MHACA introduce the agent token and reach the linear complexity, achieving lower computation cost than vanilla self-attention and cross-attention and obtain competitive performances. Our TransRFIR achieves state-of-the-art performances compared with other counterparts and is proven as an effective basic structure for image restoration. We release our project at https://github.com/GuanRunwei/FIR-CP.

{"title":"Referring flexible image restoration","authors":"Runwei Guan , Rongsheng Hu , Zhuhao Zhou , Tianlang Xue , Ka Lok Man , Jeremy Smith , Eng Gee Lim , Weiping Ding , Yutao Yue","doi":"10.1016/j.eswa.2025.126857","DOIUrl":"10.1016/j.eswa.2025.126857","url":null,"abstract":"<div><div>In reality, images often exhibit multiple degradations, such as rain and fog at night (triple degradations). However, in many cases, individuals may not want to remove all degradations, for instance, a blurry lens revealing a beautiful snowy landscape (double degradations). In such scenarios, people may only desire to deblur. These situations and requirements shed light on a new challenge in image restoration, where a model must perceive and remove specific degradation types specified by human commands in images with multiple degradations. We term this task Referring Flexible Image Restoration (RFIR). To address this, we first construct a large-scale synthetic dataset called RFIR, comprising 153,423 samples with the degraded image, text prompt for specific degradation removal and restored image. RFIR consists of five basic degradation types: blur, rain, haze, low light and snow while six main sub-categories are included for varying degrees of degradation removal. To tackle the challenge, we propose a novel transformer-based multi-task model named TransRFIR, which simultaneously perceives degradation types in the degraded image and removes specific degradation upon text prompt. TransRFIR is based on two devised modules, Multi-Head Agent Self-Attention (MHASA) for multi-degradation context modeling and Multi-Head Agent Cross Attention (MHACA) for efficient alignment between prompt and referred degradations, where MHASA and MHACA introduce the agent token and reach the linear complexity, achieving lower computation cost than vanilla self-attention and cross-attention and obtain competitive performances. Our TransRFIR achieves state-of-the-art performances compared with other counterparts and is proven as an effective basic structure for image restoration. We release our project at <span><span>https://github.com/GuanRunwei/FIR-CP</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"274 ","pages":"Article 126857"},"PeriodicalIF":7.5,"publicationDate":"2025-02-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143488618","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Learning states enhanced Knowledge Tracing: Simulating the diversity in real-world learning process

IF 7.5 1区计算机科学 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Expert Systems with Applications

Pub Date : 2025-02-23 DOI: 10.1016/j.eswa.2025.126838

Shanshan Wang , Xueying Zhang , Xun Yang , Xingyi Zhang , Keyang Wang

The Knowledge Tracing (KT) task focuses on predicting a learner’s future performance based on the historical interactions. The knowledge state plays a key role in learning process. However, considering that the knowledge state is influenced by various learning factors in the interaction process, such as the exercises similarities, responses reliability and the learner’s learning state. Previous models still face two major limitations. First, due to the exercises differences caused by various complex reasons and the unreliability of responses caused by guessing behavior, it is hard to locate the historical interaction which is most relevant to the current answered exercise. Second, the learning state is also a key factor to influence the knowledge state, which is always ignored by previous methods. To address these issues, we propose a new method named Learning State Enhanced Knowledge Tracing (LSKT). Firstly, to simulate the potential differences in interactions, inspired by Item Response Theory (IRT) paradigm, we designed three different embedding methods ranging from coarse-grained to fine-grained views and conduct comparative analysis on them. Secondly, we design a learning state extraction module to capture the changing learning state during the learning process of the learner. In turn, with the help of the extracted learning state, a more detailed knowledge state could be captured. Experimental results on four real-world datasets show that our LSKT method outperforms the current state-of-the-art methods. Our code is available at https://github.com/AcatI-B/LSKT.

{"title":"Learning states enhanced Knowledge Tracing: Simulating the diversity in real-world learning process","authors":"Shanshan Wang , Xueying Zhang , Xun Yang , Xingyi Zhang , Keyang Wang","doi":"10.1016/j.eswa.2025.126838","DOIUrl":"10.1016/j.eswa.2025.126838","url":null,"abstract":"<div><div>The Knowledge Tracing (KT) task focuses on predicting a learner’s future performance based on the historical interactions. The knowledge state plays a key role in learning process. However, considering that the knowledge state is influenced by various learning factors in the interaction process, such as the exercises similarities, responses reliability and the learner’s learning state. Previous models still face two major limitations. First, due to the exercises differences caused by various complex reasons and the unreliability of responses caused by guessing behavior, it is hard to locate the historical interaction which is most relevant to the current answered exercise. Second, the learning state is also a key factor to influence the knowledge state, which is always ignored by previous methods. To address these issues, we propose a new method named Learning State Enhanced Knowledge Tracing (LSKT). Firstly, to simulate the potential differences in interactions, inspired by Item Response Theory (IRT) paradigm, we designed three different embedding methods ranging from coarse-grained to fine-grained views and conduct comparative analysis on them. Secondly, we design a learning state extraction module to capture the changing learning state during the learning process of the learner. In turn, with the help of the extracted learning state, a more detailed knowledge state could be captured. Experimental results on four real-world datasets show that our LSKT method outperforms the current state-of-the-art methods. Our code is available at <span><span>https://github.com/AcatI-B/LSKT</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"274 ","pages":"Article 126838"},"PeriodicalIF":7.5,"publicationDate":"2025-02-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143480053","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}

引用次数: 0

Mamba-DDPM-BSA: Diffusion model based boundary sampling algorithm for imbalanced classification

IF 7.5 1区计算机科学 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Expert Systems with Applications

Pub Date : 2025-02-23 DOI: 10.1016/j.eswa.2025.126926

Fan Zhang , Quan Yuan , Xinhong Zhang

Data category imbalance is one of the major challenges in the field of medical image classification. This imbalance seriously affects the accuracy and reliability of the classification model, posing potential risks to doctors’ diagnosis and treatment. This paper proposes a Mamba-DDPM-BSA method to address the imbalanced classification issue of medical image. Firstly, the generative model Mamba-DDPM is designed for the synthesis of medical image samples. It utilizes Mamba’s global modeling capability and linear computational efficiency to improve the quality of generated samples by improving DDPM (Denoising Diffusion Probabilistic Model). Secondly, by oversampling training samples in boundary regions, the proposed Boundary Sampling Algorithm (BSA) enables synthesizer focuses more on decision boundary areas when fitting sample distributions. This approach generates more samples near the decision boundary, pushing the decision boundary that originally intrudes into the minority class distribution towards the true distribution. Finally, a Mamba-DDPM-BSA method is proposed, which adopts an interactive synthesis method and makes full use of diffusion generation model and Boundary Sampling Algorithm to interact with the classification model, aiming to synthesize images that target the defects of the classification model to improve the discriminative ability and robustness of the classifier. Experiments based on HAM10000 data set show that Mamba-DDPM-BSA reaches 81.03%, 82.14%, and 82.71% on Matthew’s correlation coefficient, Balanced Accuracy, and Macro F1, respectively. The proposed method is superior to the traditional imbalanced classification method.

{"title":"Mamba-DDPM-BSA: Diffusion model based boundary sampling algorithm for imbalanced classification","authors":"Fan Zhang , Quan Yuan , Xinhong Zhang","doi":"10.1016/j.eswa.2025.126926","DOIUrl":"10.1016/j.eswa.2025.126926","url":null,"abstract":"<div><div>Data category imbalance is one of the major challenges in the field of medical image classification. This imbalance seriously affects the accuracy and reliability of the classification model, posing potential risks to doctors’ diagnosis and treatment. This paper proposes a Mamba-DDPM-BSA method to address the imbalanced classification issue of medical image. Firstly, the generative model Mamba-DDPM is designed for the synthesis of medical image samples. It utilizes Mamba’s global modeling capability and linear computational efficiency to improve the quality of generated samples by improving DDPM (Denoising Diffusion Probabilistic Model). Secondly, by oversampling training samples in boundary regions, the proposed Boundary Sampling Algorithm (BSA) enables synthesizer focuses more on decision boundary areas when fitting sample distributions. This approach generates more samples near the decision boundary, pushing the decision boundary that originally intrudes into the minority class distribution towards the true distribution. Finally, a Mamba-DDPM-BSA method is proposed, which adopts an interactive synthesis method and makes full use of diffusion generation model and Boundary Sampling Algorithm to interact with the classification model, aiming to synthesize images that target the defects of the classification model to improve the discriminative ability and robustness of the classifier. Experiments based on HAM10000 data set show that Mamba-DDPM-BSA reaches 81.03%, 82.14%, and 82.71% on Matthew’s correlation coefficient, Balanced Accuracy, and Macro F1, respectively. The proposed method is superior to the traditional imbalanced classification method.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"274 ","pages":"Article 126926"},"PeriodicalIF":7.5,"publicationDate":"2025-02-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143488691","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}

引用次数: 0

Selection of best location for household waste recycling plants using novel information measures and algorithm in fermatean fuzzy environment

IF 7.5 1区计算机科学 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Expert Systems with Applications

Pub Date : 2025-02-22 DOI: 10.1016/j.eswa.2025.126897

Mrinmay Pathak , Mausumi Sen , Suganya Devi K.

The selection of appropriate locations for household waste recycling plants is a critical issue due to its impact on environmental sustainability and waste management efficiency. By finding the best spot, we can make recycling more efficient, reduce pollution, and improve sustainability. This study presents innovative methods within the Fermatean fuzzy sets (FFSs) specifically, we introduce a novel distance measure, a logarithmic score function, an entropy measure and using this measures we introduced an integrated method combining Entropy Measure (EM), Step-wise Weight Assessment Ratio Analysis (SWARA), and Weighted Aggregated Sum Product Assessment (WASPAS) methods for this purpose. Fermatean fuzzy sets provide a broader range of membership, nonmembership, and hesitation values, offering greater flexibility and expressiveness in evaluating alternatives. Our proposed measures and the algorithm are compared with existing measures and the methods in the FFSs environment, demonstrating their reliability and superiority by addressing the limitations of existing measures and methodologies. This combined framework leverages both subjective (by SWARA) and objective (by Entropy Measure) weighting to comprehensively evaluate alternatives against multiple criteria and experts which is free from biases towards one particular weight model. The innovative distance measure and score function significantly improve decision-making reliability with the help of our proposed algorithm, especially under conditions of uncertainty. Experimental results and a real-life case study confirm that our approach not only provides finer differentiation than existing fuzzy techniques but also enhances the robustness and accuracy of MCDM outcomes. This advancement is particularly valuable in the context of selecting optimal locations for household waste recycling plants, contributing to more effective and sustainable waste management solutions.

{"title":"Selection of best location for household waste recycling plants using novel information measures and algorithm in fermatean fuzzy environment","authors":"Mrinmay Pathak , Mausumi Sen , Suganya Devi K.","doi":"10.1016/j.eswa.2025.126897","DOIUrl":"10.1016/j.eswa.2025.126897","url":null,"abstract":"<div><div>The selection of appropriate locations for household waste recycling plants is a critical issue due to its impact on environmental sustainability and waste management efficiency. By finding the best spot, we can make recycling more efficient, reduce pollution, and improve sustainability. This study presents innovative methods within the Fermatean fuzzy sets (FFSs) specifically, we introduce a novel distance measure, a logarithmic score function, an entropy measure and using this measures we introduced an integrated method combining Entropy Measure (EM), Step-wise Weight Assessment Ratio Analysis (SWARA), and Weighted Aggregated Sum Product Assessment (WASPAS) methods for this purpose. Fermatean fuzzy sets provide a broader range of membership, nonmembership, and hesitation values, offering greater flexibility and expressiveness in evaluating alternatives. Our proposed measures and the algorithm are compared with existing measures and the methods in the FFSs environment, demonstrating their reliability and superiority by addressing the limitations of existing measures and methodologies. This combined framework leverages both subjective (by SWARA) and objective (by Entropy Measure) weighting to comprehensively evaluate alternatives against multiple criteria and experts which is free from biases towards one particular weight model. The innovative distance measure and score function significantly improve decision-making reliability with the help of our proposed algorithm, especially under conditions of uncertainty. Experimental results and a real-life case study confirm that our approach not only provides finer differentiation than existing fuzzy techniques but also enhances the robustness and accuracy of MCDM outcomes. This advancement is particularly valuable in the context of selecting optimal locations for household waste recycling plants, contributing to more effective and sustainable waste management solutions.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"274 ","pages":"Article 126897"},"PeriodicalIF":7.5,"publicationDate":"2025-02-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143474349","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}

引用次数: 0