Applied Soft Computing最新文献

筛选
英文 中文
Interval-valued Pythagorean fuzzy distance-based extended inferior ratio method for multiattribute decision-making: Application to green supplier selection in manufacturing industry 基于区间值毕达哥拉斯模糊距离的多属性决策扩展劣比法:在制造业绿色供应商选择中的应用
IF 6.6 1区 计算机科学
Applied Soft Computing Pub Date : 2025-09-18 DOI: 10.1016/j.asoc.2025.113935
Zhe Liu , Donglai Wang , Muhammet Deveci , Sukumar Letchmunan
{"title":"Interval-valued Pythagorean fuzzy distance-based extended inferior ratio method for multiattribute decision-making: Application to green supplier selection in manufacturing industry","authors":"Zhe Liu ,&nbsp;Donglai Wang ,&nbsp;Muhammet Deveci ,&nbsp;Sukumar Letchmunan","doi":"10.1016/j.asoc.2025.113935","DOIUrl":"10.1016/j.asoc.2025.113935","url":null,"abstract":"<div><div>Interval-valued Pythagorean fuzzy sets (IVPFSs) have emerged as a powerful tool for handling uncertainty and vagueness in multiattribute decision-making (MADM). In this paper, we first propose a novel distance measure for IVPFSs based on triangular divergence, which satisfies all core distance axioms and significantly improves discrimination ability compared to existing measures. Building on this, we introduce a maximizing deviation strategy with a new loss function to objectively determine attribute weights. Furthermore, we develop an extended inferior ratio (EIR) method that incorporates a dynamic weight parameter to flexibly balance the influence of positive and negative ideal solutions. The performance of the proposed method is demonstrated through a case study on green supplier selection in the manufacturing industry. The results indicate that, among the seven criteria evaluated, the most suitable suppliers are ranked as follows: <span><math><mi>β</mi></math></span> (1.0000), <span><math><mi>α</mi></math></span> (0.6471), <span><math><mi>δ</mi></math></span> (0.3500), <span><math><mi>ϵ</mi></math></span> (0.0690), and <span><math><mi>θ</mi></math></span> (0.0000). In addition, sensitivity and comparative analyses confirm the robustness and consistency of the proposed method, reflecting its effectiveness and practical value for sustainable decision-making in real-world scenarios.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"185 ","pages":"Article 113935"},"PeriodicalIF":6.6,"publicationDate":"2025-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145159715","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
Deep cross-visual semantic hashing with self-calibrated collaborative attention 基于自校准协同注意的深度跨视觉语义哈希
IF 6.6 1区 计算机科学
Applied Soft Computing Pub Date : 2025-09-18 DOI: 10.1016/j.asoc.2025.113937
Hao Feng, Xiangbo Zhou, Yue Wu, Jian Zhou, Banglei Zhao
{"title":"Deep cross-visual semantic hashing with self-calibrated collaborative attention","authors":"Hao Feng,&nbsp;Xiangbo Zhou,&nbsp;Yue Wu,&nbsp;Jian Zhou,&nbsp;Banglei Zhao","doi":"10.1016/j.asoc.2025.113937","DOIUrl":"10.1016/j.asoc.2025.113937","url":null,"abstract":"<div><div>Deep hashing has garnered considerable attention due to its remarkable retrieval efficiency and low storage cost, particularly in visual retrieval scenarios. However, current deep hashing methods generally integrate hash coding into a single-stream architecture, which limits the discriminative power of learned visual features and yields suboptimal hash codes. Additionally, over-reliance on semantic labels shared across samples fails to fully exploit the intrinsic semantic correlations between labels and corresponding visual features. To address these issues, we propose a deep cross-visual semantic hashing (DCvSH) method for image retrieval. First, we develop a visual image feature decoupling encoding network that leverages a self-calibrated collaborative attention mechanism to disentangle common and specific semantics across related images. These decoupled features are fed into a shared decoder for image reconstruction, yielding discriminative visual feature representations. Second, we construct a cross-visual semantic representation learning network with a two-level multi-layer perceptron to capture the underlying relationships between semantic label encodings and visual feature embeddings, while a hypergraph structure is introduced to preserve pairwise similarity relationships. Experimental results on the CIFAR-10, NUS-WIDE, and MIRFLICKR datasets demonstrate consistent improvements, with average mean average precision (mAP) scores reaching 0.895, 0.874, and 0.881 at different code lengths, respectively. Notably, DCvSH outperforms other baselines across all evaluation metrics.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"185 ","pages":"Article 113937"},"PeriodicalIF":6.6,"publicationDate":"2025-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145119772","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
Fusing local density and approximate distance for nonparametric outlier detection 融合局部密度和近似距离的非参数离群点检测
IF 6.6 1区 计算机科学
Applied Soft Computing Pub Date : 2025-09-18 DOI: 10.1016/j.asoc.2025.113898
Zhiyu Chen , Can Gao , Jie Zhou , Ying Yu
{"title":"Fusing local density and approximate distance for nonparametric outlier detection","authors":"Zhiyu Chen ,&nbsp;Can Gao ,&nbsp;Jie Zhou ,&nbsp;Ying Yu","doi":"10.1016/j.asoc.2025.113898","DOIUrl":"10.1016/j.asoc.2025.113898","url":null,"abstract":"<div><div>Outlier detection is an essential yet challenging task in intelligent data analysis, and some density-based unsupervised methods have been introduced to identify outliers in low-density regions. However, these methods still suffer from inaccurate density estimation and limited capability in detecting diverse types of outliers. In this study, we propose a nonparametric outlier detection method with the fusion of density and distance (POD-FDD). The proposed method employs adaptive kernel density estimation based on natural neighborhoods, which reduces the sensitivity to parameters in density estimation. Moreover, the optimistic and pessimistic densities are introduced to enhance the reliability of density estimation in the local neighborhood. In addition, approximate reachability distance information is integrated to improve the capability of identifying cluster outliers. Ultimately, a robust parametric-free outlier detection method is developed to detect different types of outliers. Extensive comparative experiments and statistical significance analysis on synthetic and public datasets demonstrate its superior performance, achieving an average improvement of 1.97 % in the AUC metric.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"185 ","pages":"Article 113898"},"PeriodicalIF":6.6,"publicationDate":"2025-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145158921","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
Extracting knowledge from limited data: An updated review of data-driven and model-driven few-shot learning for agriculture 从有限的数据中提取知识:数据驱动和模型驱动的农业短时间学习的最新综述
IF 6.6 1区 计算机科学
Applied Soft Computing Pub Date : 2025-09-18 DOI: 10.1016/j.asoc.2025.113968
Kam Meng Goh , Usman Ullah Sheikh , Jun Kit Chaw , Weng Kin Lai , Weng Chun Tan , Santhi Krishnamoorthy
{"title":"Extracting knowledge from limited data: An updated review of data-driven and model-driven few-shot learning for agriculture","authors":"Kam Meng Goh ,&nbsp;Usman Ullah Sheikh ,&nbsp;Jun Kit Chaw ,&nbsp;Weng Kin Lai ,&nbsp;Weng Chun Tan ,&nbsp;Santhi Krishnamoorthy","doi":"10.1016/j.asoc.2025.113968","DOIUrl":"10.1016/j.asoc.2025.113968","url":null,"abstract":"<div><div>Deep learning has demonstrated considerable success in agricultural applications. However, its conventional implementations heavily depend on large-scale labelled datasets—a requirement that is often impractical in agriculture due to data scarcity, high annotation costs, or environmental variability. While insufficient training data can significantly limit the performance of standard deep learning models, Few-Shot Learning (FSL) has emerged as a transformative paradigm, enabling robust model training with minimal labelled samples by utilising limited data for training instead. Despite its potential, a critical review assessing how FSL addresses expert system challenges in agriculture remains notably absent. This paper attempts to fill this void by presenting an updated comprehensive review of FSL's applications in agriculture. We categorise FSL methodologies into two primary approaches: data processing-driven and model learning-driven. Data processing–driven approaches address data scarcity by enriching representational diversity through synthetic samples generated with models such as generative adversarial networks, or by transferring knowledge from related domains to improve generalisation. In contrast, model learning–driven strategies confront the same challenge through specialised architectures and optimisation techniques that enable effective generalisation from limited samples. Within this taxonomy, data processing–driven paradigms include transfer learning and generative artificial intelligence, while model learning–driven paradigms cover metric learning methods such as Siamese or prototypical networks, together with model-based and optimisation approaches designed for efficient generalisation. Our analysis pinpoints cutting-edge technologies within each sector, shedding light on overlooked areas and opportunities where FSL can harness limited data to yield promising outcomes when used to solve problems in agriculture.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"185 ","pages":"Article 113968"},"PeriodicalIF":6.6,"publicationDate":"2025-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145159811","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
DefGCL: Defence-enhanced graph contrastive learning against attribute inference attacks DefGCL:针对属性推理攻击的防御增强图对比学习
IF 6.6 1区 计算机科学
Applied Soft Computing Pub Date : 2025-09-18 DOI: 10.1016/j.asoc.2025.113911
Jinyin Chen , Fanyu Ao , Wenbo Mu , Haiyang Xiong
{"title":"DefGCL: Defence-enhanced graph contrastive learning against attribute inference attacks","authors":"Jinyin Chen ,&nbsp;Fanyu Ao ,&nbsp;Wenbo Mu ,&nbsp;Haiyang Xiong","doi":"10.1016/j.asoc.2025.113911","DOIUrl":"10.1016/j.asoc.2025.113911","url":null,"abstract":"<div><div>Graph-structured data are prevalent in many real-world applications, such as social networks, drug discovery, and fraud detection. While Graph Neural Networks (GNNs) have shown remarkable performance by capturing rich relational patterns, their success often relies on large labeled datasets and raises growing privacy concerns. Graph Contrastive Learning (GCL) has emerged as a powerful unsupervised alternative by leveraging data augmentations to learn robust representations without labeled data. However, recent studies reveal that GCL models are particularly vulnerable to attribute inference attacks, and existing works prioritize performance improvement over privacy protection. To address this issue, we propose a <u>D</u><u>e</u><u>f</u>ense-enhanced <u>G</u>raph <u>C</u>ontrastive <u>L</u>earning, dubbed <em>DefGCL</em>, that integrates four coordinated defense strategies to enhance privacy without degrading utility. Specifically, DefGCL employs edge-based graph augmentations to limit exposure to structural attributes, selects negative samples with low attribute sensitivity scores to reduce leakage, modifies the contrastive loss to decouple graph embeddings from attributes, and injects differential privacy noise during the embedding stage. Extensive experiments on five benchmark datasets demonstrate that DefGCL achieves state-of-the-art (SOTA) performance in both privacy preservation and task accuracy. For instance, on the AIDS dataset, DefGCL reduces attribute inference accuracy by 35 % while incurring only a 0.60 % drop in main task performance. Additionally, DefGCL improves computational efficiency by reducing runtime by nearly 50 % compared to baseline methods.<span><span><sup>1</sup></span></span></div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"185 ","pages":"Article 113911"},"PeriodicalIF":6.6,"publicationDate":"2025-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145119680","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
Extended belief rule-based system with online joint learning strategy 基于扩展信念规则的在线联合学习系统
IF 6.6 1区 计算机科学
Applied Soft Computing Pub Date : 2025-09-18 DOI: 10.1016/j.asoc.2025.113901
Bingbing Hou , Min Xue , Leilei Chang , Zijian Wu
{"title":"Extended belief rule-based system with online joint learning strategy","authors":"Bingbing Hou ,&nbsp;Min Xue ,&nbsp;Leilei Chang ,&nbsp;Zijian Wu","doi":"10.1016/j.asoc.2025.113901","DOIUrl":"10.1016/j.asoc.2025.113901","url":null,"abstract":"<div><div>With the proliferation of dynamic data streams in the advanced technology environment, it is necessary to solve the evolving classification problems by using adaptable and interpretable artificial intelligence techniques. To meet this challenge, a new extended belief rule-based (EBRB) system incorporating online joint learning strategy is proposed in this paper. The online joint learning strategy comprises two key components: rule update and parameter update schemes. In the rule update scheme, different rule incorporation processes are designed for the labeled or unlabeled input data while overlapping and redundant rules are removed from the rule base. To adapt the updated rule base, the parameter update scheme is designed to retune the parameters within updated rule base. The antecedent attribute weights are optimized using the Bayesian optimization algorithm and the rule weights are updated based on the consistency of rules. To evaluate the performance of the developed system, it is applied to assist radiologists in diagnosing thyroid nodules. Compared with the existing offline EBRB systems and online learning methods, the proposed online joint learning EBRB system could generate higher classification accuracy with fewer rules in the limited running time.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"185 ","pages":"Article 113901"},"PeriodicalIF":6.6,"publicationDate":"2025-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145159664","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
WamGLM: A multimodal large-scale language model for wafer map defect information in-depth query through multi-turn dialogue based on prototypical supervised contrastive learning WamGLM:基于原型监督对比学习的多回合对话深度查询晶圆图缺陷信息的多模态大规模语言模型
IF 6.6 1区 计算机科学
Applied Soft Computing Pub Date : 2025-09-18 DOI: 10.1016/j.asoc.2025.113962
Shulong Gu , Zihao Lei , Guangrui Wen , Quanning Xu , Zhaojun Steven Li , Xuefeng Chen , Chunsheng Yang
{"title":"WamGLM: A multimodal large-scale language model for wafer map defect information in-depth query through multi-turn dialogue based on prototypical supervised contrastive learning","authors":"Shulong Gu ,&nbsp;Zihao Lei ,&nbsp;Guangrui Wen ,&nbsp;Quanning Xu ,&nbsp;Zhaojun Steven Li ,&nbsp;Xuefeng Chen ,&nbsp;Chunsheng Yang","doi":"10.1016/j.asoc.2025.113962","DOIUrl":"10.1016/j.asoc.2025.113962","url":null,"abstract":"<div><div>To ensure production efficiency and process stability in semiconductor manufacturing, it is of critical importance to detect wafer map defects and perform information query for tracing and solving problems during the manufacturing process. Numerous vision models based on deep learning have been successfully applied to wafer map defect recognition (WMDR), yielding remarkable results. However, the dynamic and in-depth querying of wafer map defect information remains relatively underexplored. Leveraging the rapid advancements in multimodal large language models (MLLMs), this paper proposes a novel approach for wafer map defect information query (WMDIQ). First, following the paradigm of employing cross-modal alignment model to bridge vision and language models, an end-to-end response MLLM: general language model for wafer map (WamGLM), is constructed for WMDIQ. Concurrently, by designing an interactive dialogue framework between large language models (LLMs), the first large-scale multi-turn dialogue dataset: visual multi-turn question answering dataset for wafer map defects (WaferMapVMQA Dataset), is constructed for wafer map defect analysis. Subsequently, WamGLM is trained using a two-stage fine-tuning strategy. In the first stage, a visual fine-tuning method based on prototypical supervised contrastive learning (PSCL) is introduced to enhance the intra-class compactness and inter-class separability of defect features. In the second stage, language fine-tuning is conducted using the WaferMapVMQA Dataset to infuse specialized knowledge into WamGLM. To validate the effectiveness and superiority of the proposed method, experiments are conducted on a real wafer map dataset. The results demonstrate that the proposed method significantly outperforms other methods in both defect recognition performance and information query response performance. Our code is available at: <span><span>https://github.com/ZihaoLei/WamGLM</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"185 ","pages":"Article 113962"},"PeriodicalIF":6.6,"publicationDate":"2025-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145159667","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
Adaptive fuzzy entropy optimization with opposition-based archimedes search for robust multilevel image segmentation 基于阿基米德搜索的自适应模糊熵优化鲁棒多级图像分割
IF 6.6 1区 计算机科学
Applied Soft Computing Pub Date : 2025-09-18 DOI: 10.1016/j.asoc.2025.113943
Anusha Ganesan , Sungho Kim , Ganesan Nagabushnam
{"title":"Adaptive fuzzy entropy optimization with opposition-based archimedes search for robust multilevel image segmentation","authors":"Anusha Ganesan ,&nbsp;Sungho Kim ,&nbsp;Ganesan Nagabushnam","doi":"10.1016/j.asoc.2025.113943","DOIUrl":"10.1016/j.asoc.2025.113943","url":null,"abstract":"<div><div>Image segmentation plays a critical role in diverse computer vision applications. Multilevel thresholding (MLT) remains one of its most widely used unsupervised techniques due to its simplicity and interpretability. However, existing MLT methods often suffer from two major limitations: (1) the inability to adapt to local intensity variations and (2) the computational burden associated with high-dimensional threshold search. To address these challenges, this study proposes a novel segmentation framework that integrates a Proximity-Adaptive Fuzzy Entropy (PAFE) model with an Opposition-Based Learning-enhanced Archimedes Optimization Algorithm (OBL-EAOA). The PAFE model utilizes dynamically adjusted trapezoidal membership functions based on intensity proximity to candidate thresholds, allowing for a more adaptive and smooth entropy surface. Meanwhile, the OBL-EAOA enhances optimization performance through opposition-based learning and adaptive parameter control, improving exploration diversity and convergence speed. The proposed PAFE-EAOA framework is validated on two benchmark datasets, BSD500 and PASCAL VOC 2012, using five standard metrics: PSNR, SSIM, FSIM, SNR, and computation time. Compared with several state-of-the-art methods including Kapur Entropy (KE)-EAOA, Fuzzy Entropy (FE)-EAOA, Patch-Levy-Based Bees Algorithm (PLBA), Marine Predators Algorithm (MPA), Improved Grey Wolf Optimizer (IGWO), and standard Archimedes Optimization Algorithm (AOA), the proposed approach consistently achieves superior segmentation quality. Notably, it reduces computation time by up to 60 % and achieves statistically significant improvements, as confirmed by the Wilcoxon signed-rank test. These results demonstrate the framework’s robustness, scalability, and effectiveness for real-world MLT-based image segmentation.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"185 ","pages":"Article 113943"},"PeriodicalIF":6.6,"publicationDate":"2025-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145158925","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
Multi-objective genetic programming for binary classification with adaptive thresholds and a generalization-optimizing fitness function 具有自适应阈值和广义优化适应度函数的二值分类多目标遗传规划
IF 6.6 1区 计算机科学
Applied Soft Computing Pub Date : 2025-09-18 DOI: 10.1016/j.asoc.2025.113956
Minghui Bai , Yuan Gao , Xiaoying Gao , Jianbin Ma
{"title":"Multi-objective genetic programming for binary classification with adaptive thresholds and a generalization-optimizing fitness function","authors":"Minghui Bai ,&nbsp;Yuan Gao ,&nbsp;Xiaoying Gao ,&nbsp;Jianbin Ma","doi":"10.1016/j.asoc.2025.113956","DOIUrl":"10.1016/j.asoc.2025.113956","url":null,"abstract":"<div><div>Genetic programming (GP) has been widely applied to classifier construction due to its flexible representation and powerful feature construction capabilities. Existing studies have proposed various fitness functions to improve GP-based classifiers, but most of them rely on a fixed decision threshold. However, when dealing with imbalanced classification problems, a fixed threshold often biases the model toward the majority class, thereby compromising overall performance. To address this issue, in this paper, we propose a novel multi-objective GP framework for constructing binary classifiers with adaptive threshold adjustment. During evolution, the method employs Youden’s Index to dynamically adjust the threshold of each individual, enabling the classifiers to better fit the underlying data distribution. In addition, we introduce a new class separation metric, <em>dist</em><sub>t</sub>, to quantify the clarity of class boundaries and enhance the generalization ability of the evolved models. The framework jointly optimizes three objectives: minority class accuracy, majority class accuracy, and the proposed <em>dist</em><sub>t</sub> metric. Experiments on 14 imbalanced datasets demonstrate that our method significantly outperforms conventional single-objective GP with fixed thresholds. Further results also confirm the positive impact of the proposed <em>dist</em><sub>t</sub> metric on classification performance. Compared to seven existing GP algorithms and five traditional machine learning classifiers, our approach achieves superior overall performance and better generalization ability.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"185 ","pages":"Article 113956"},"PeriodicalIF":6.6,"publicationDate":"2025-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145159669","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
Meta-learning-based adaptive operator selection for traveling salesman problem 基于元学习的旅行商问题自适应算子选择
IF 6.6 1区 计算机科学
Applied Soft Computing Pub Date : 2025-09-17 DOI: 10.1016/j.asoc.2025.113930
Ho Young Jeong , Byung Duk Song
{"title":"Meta-learning-based adaptive operator selection for traveling salesman problem","authors":"Ho Young Jeong ,&nbsp;Byung Duk Song","doi":"10.1016/j.asoc.2025.113930","DOIUrl":"10.1016/j.asoc.2025.113930","url":null,"abstract":"<div><div>In evolutionary optimization, effectively leveraging knowledge about search operator performance is crucial for enhancing algorithmic results. Traditional operator selection strategies often rely on fixed heuristics or trial-and-error, which struggle to adapt to the nonstationary search dynamics of evolutionary runs—i.e., the stage-dependent, instance-dependent, and population-dependent shifts in operator effectiveness—and typically yield suboptimal performance. To address these challenges, we propose a novel meta-learning-based adaptive operator selection (AOS) framework. It leverages a Long Short-Term Memory (LSTM) neural network to learn temporal patterns of operator performance from historical data and dynamically adjust operator choice on-the-fly. The framework also integrates domain-specific biases to preserve population diversity and promote effective exploration, and it continuously updates its selection policy through dynamic online learning as the evolutionary process unfolds. Experiments on the Traveling Salesman Problem (TSP) benchmark demonstrate that the proposed LSTM-based AOS method significantly outperforms conventional approaches to operator selection. In particular, it achieved a median optimality gap of 9.87 % on a suite of TSP instances—approximately a 20 % improvement over the best fixed-operator configuration—indicating superior solution quality. Moreover, our approach consistently surpassed other state-of-the-art AOS techniques, underscoring the efficacy of the LSTM-driven framework and its significant potential to enhance evolutionary algorithm performance on complex optimization tasks.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"185 ","pages":"Article 113930"},"PeriodicalIF":6.6,"publicationDate":"2025-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145158919","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
0
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
相关产品
×
本文献相关产品
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:604180095
Book学术官方微信