首页 > 最新文献
paperpass

Applied Soft Computing最新文献

英文 中文
Automated detection of tomato ripeness in natural environment using a coupled deep learning model of Sequencer and MLP 基于Sequencer和MLP耦合深度学习模型的自然环境下番茄成熟度自动检测
IF 7.8 1区 计算机科学 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE Pub Date : 2026-07-21 DOI: 10.1016/j.asoc.2026.116038
Faying Wu, Mingxuan Li, Chongda Zhou, Feifan Zhang
Accurate tomato ripeness detection is crucial for automated harvesting, yet remains challenging under complex field conditions such as uneven illumination, occlusion, and background clutter. This paper proposes a lightweight coupled framework, Sequencer+MLP, where the attention modules are replaced by 2D BiLSTM-based Sequencer blocks to efficiently capture spatial dependencies, and a compact MLP head performs final classification. Experiments were conducted on a greenhouse dataset collected under diverse lighting and occlusion scenarios. The proposed method achieved 96.8% precision, 94.8% recall, and 98.9% classification mAP, while maintaining an inference latency of 0.6135 s with a model size of 2.4 MB, indicating potential for practical deployment after further edge-device validation. Additional evaluations under illumination and shelter subsets further verified robustness, with mAP above 95%. These results demonstrate that the coupled Sequencer+MLP design provides an effective and efficient solution for tomato ripeness detection in natural agricultural environments. The MLP classifier achieved the best classification performance among the tested heads, indicating stable behavior across repeated trials.
准确的番茄成熟度检测对自动化收获至关重要,但在复杂的野外条件下,如光照不均匀、遮挡和背景杂波,仍然具有挑战性。本文提出了一种轻量级的耦合框架Sequencer+MLP,其中注意力模块被基于二维bilstm的Sequencer块取代,以有效捕获空间依赖性,紧凑的MLP头执行最终分类。在不同光照和遮挡条件下的温室数据集上进行了实验。该方法的准确率为96.8%,召回率为94.8%,分类mAP为98.9%,同时在模型大小为2.4 MB的情况下保持了0.6135 s的推理延迟,表明在进一步的边缘设备验证后具有实际部署的潜力。在照明和遮蔽子集下的附加评估进一步验证了稳健性,mAP超过95%。这些结果表明,Sequencer+MLP耦合设计为自然农业环境下的番茄成熟度检测提供了一种有效的解决方案。MLP分类器在被试头部中获得了最好的分类性能,在重复试验中表现出稳定的行为。
{"title":"Automated detection of tomato ripeness in natural environment using a coupled deep learning model of Sequencer and MLP","authors":"Faying Wu,&nbsp;Mingxuan Li,&nbsp;Chongda Zhou,&nbsp;Feifan Zhang","doi":"10.1016/j.asoc.2026.116038","DOIUrl":"10.1016/j.asoc.2026.116038","url":null,"abstract":"<div><div>Accurate tomato ripeness detection is crucial for automated harvesting, yet remains challenging under complex field conditions such as uneven illumination, occlusion, and background clutter. This paper proposes a lightweight coupled framework, Sequencer+MLP, where the attention modules are replaced by 2D BiLSTM-based Sequencer blocks to efficiently capture spatial dependencies, and a compact MLP head performs final classification. Experiments were conducted on a greenhouse dataset collected under diverse lighting and occlusion scenarios. The proposed method achieved 96.8% precision, 94.8% recall, and 98.9% classification mAP, while maintaining an inference latency of 0.6135<!--> <!-->s with a model size of 2.4 MB, indicating potential for practical deployment after further edge-device validation. Additional evaluations under illumination and shelter subsets further verified robustness, with mAP above 95%. These results demonstrate that the coupled Sequencer+MLP design provides an effective and efficient solution for tomato ripeness detection in natural agricultural environments. The MLP classifier achieved the best classification performance among the tested heads, indicating stable behavior across repeated trials.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"203 ","pages":"Article 116038"},"PeriodicalIF":7.8,"publicationDate":"2026-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148575919","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
Conditional generation-augmented multi-source domain generalization for rolling bearing RUL prediction via manifold learning-based health indicator construction and adaptive ensemble learning 基于流形学习健康指标构建和自适应集成学习的条件生成增强多源域泛化滚动轴承RUL预测
IF 7.8 1区 计算机科学 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE Pub Date : 2026-07-19 DOI: 10.1016/j.asoc.2026.116003
Jimeng Li, Haobo Zhang, Hongliang Sun, Jinfeng Zhang
Predictive maintenance of equipment relies on the accurate prediction of remaining useful life (RUL); however, fluctuating operating conditions in industrial scenarios often cause significant distribution shifts between training and testing data. Domain generalization (DG) can mitigate such domain shifts without requiring access to target domain test data. However, negative transfer caused by multi-source domain knowledge conflicts and inappropriate knowledge fusion has long been a key factor restricting the generalization and prediction accuracy of the model. To address these challenges, this paper proposes a novel DG framework for RUL prediction based on conditional generation-based augmentation. First, a robust health indicator is constructed via manifold learning, and a piecewise function based on degradation rate is designed to encode RUL labels. Second, a conditioned diffusion model incorporating hybrid source domain constraints is developed to generate task-relevant synthetic degradation-feature samples using source domain information, thereby enriching the training distribution and improving robustness to unseen operating-condition shifts. Third, a TimeSerFormer network integrating temporal convolutional networks and Transformer architectures within a main-auxiliary task design is proposed to capture multi-scale degradation features. Additionally, dual-distribution alignment and adaptive ensemble learning are employed to mitigate negative transfer across source domains. Experimental results on the IEEE PHM 2012 and XJTU-SY bearing datasets demonstrate that the proposed method achieves mean RMSE and mean MAE of 0.1248 and 0.0964, respectively, which outperforms several state-of-the-art DG methods and provides a promising solution for reliable cross-condition RUL prediction of rolling bearings.
设备的预测性维护依赖于对剩余使用寿命(RUL)的准确预测;然而,工业场景中波动的操作条件通常会导致训练和测试数据之间的显著分布变化。领域泛化(DG)可以在不需要访问目标领域测试数据的情况下减轻这种领域转移。然而,多源领域知识冲突和知识融合不当导致的负迁移一直是制约模型泛化和预测精度的关键因素。为了解决这些挑战,本文提出了一种基于条件生成增强的规则生成预测DG框架。首先,通过流形学习构造鲁棒健康指标,并设计基于退化率的分段函数对规则标签进行编码。其次,建立了一个包含混合源域约束的条件扩散模型,利用源域信息生成与任务相关的合成退化特征样本,从而丰富了训练分布,提高了对未知操作条件变化的鲁棒性。第三,提出了在主辅助任务设计中集成时间卷积网络和Transformer架构的TimeSerFormer网络,以捕获多尺度退化特征。此外,采用双分布对齐和自适应集成学习来减轻源域间的负迁移。在IEEE PHM 2012和XJTU-SY轴承数据集上的实验结果表明,该方法的平均RMSE和平均MAE分别为0.1248和0.0964,优于几种最先进的DG方法,为可靠的滚动轴承交叉条件RUL预测提供了一种有希望的解决方案。
{"title":"Conditional generation-augmented multi-source domain generalization for rolling bearing RUL prediction via manifold learning-based health indicator construction and adaptive ensemble learning","authors":"Jimeng Li,&nbsp;Haobo Zhang,&nbsp;Hongliang Sun,&nbsp;Jinfeng Zhang","doi":"10.1016/j.asoc.2026.116003","DOIUrl":"10.1016/j.asoc.2026.116003","url":null,"abstract":"<div><div>Predictive maintenance of equipment relies on the accurate prediction of remaining useful life (RUL); however, fluctuating operating conditions in industrial scenarios often cause significant distribution shifts between training and testing data. Domain generalization (DG) can mitigate such domain shifts without requiring access to target domain test data. However, negative transfer caused by multi-source domain knowledge conflicts and inappropriate knowledge fusion has long been a key factor restricting the generalization and prediction accuracy of the model. To address these challenges, this paper proposes a novel DG framework for RUL prediction based on conditional generation-based augmentation. First, a robust health indicator is constructed via manifold learning, and a piecewise function based on degradation rate is designed to encode RUL labels. Second, a conditioned diffusion model incorporating hybrid source domain constraints is developed to generate task-relevant synthetic degradation-feature samples using source domain information, thereby enriching the training distribution and improving robustness to unseen operating-condition shifts. Third, a TimeSerFormer network integrating temporal convolutional networks and Transformer architectures within a main-auxiliary task design is proposed to capture multi-scale degradation features. Additionally, dual-distribution alignment and adaptive ensemble learning are employed to mitigate negative transfer across source domains. Experimental results on the IEEE PHM 2012 and XJTU-SY bearing datasets demonstrate that the proposed method achieves mean RMSE and mean MAE of 0.1248 and 0.0964, respectively, which outperforms several state-of-the-art DG methods and provides a promising solution for reliable cross-condition RUL prediction of rolling bearings.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"203 ","pages":"Article 116003"},"PeriodicalIF":7.8,"publicationDate":"2026-07-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148575917","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
Emotion-Based Reinforcement Learning for Passivity-Based Control of Redundant Manipulators in Challenging Environments 基于情绪的冗余机械臂被动控制强化学习
IF 7.8 1区 计算机科学 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE Pub Date : 2026-07-19 DOI: 10.1016/j.asoc.2026.115997
Rami Al-Khulaidi, Tien-Fu Lu, Steven Grainger, Rini Akmeliawati
Controlling redundant manipulators in unstructured environments like agricultural settings is challenging due to high nonlinearities and the need for both precision and stability. This paper presents a novel intelligent control architecture that integrates a stable-by-design Passivity-Based Control (PBC) law with a bio-inspired reinforcement learning tuner. The core novelty lies in our Human Learning Behaviour (HLB) model, which, unlike conventional reward shaping, employs a finite-state machine to dynamically modulate the exploration strategy of a Deep Deterministic Policy Gradient (DDPG) agent. This state machine transitions between discrete states, framed as artificial emotions (e.g., Joy, Fear), based on real-time performance indicators. Each state directly adjusts the agent's exploration noise, enabling it to learn more efficiently by exploring aggressively when performance is poor and exploiting known good policies when performance is high. This adaptive tuner intelligently adjusts the proportional (Kp) and differential (KD) gains of the PBC law, which is derived from a Port-Controlled Hamiltonian (PCH) model to ensure the underlying manipulator dynamics remain stable for any positive-definite gains selected by the learning agent. Validated on an 8-DOF manipulator in simulated, obstacle-rich agricultural scenarios, the HLB-driven controller demonstrated a 77–80% improvement in trajectory tracking and a 30.7% reduction in energy consumption compared to fixed-gain and standard DDPG baselines, while maintaining a final tracking error of approximately 10⁻⁴ radians. The results confirm that our FSM-based adaptive exploration strategy yields a highly precise, energy-efficient, and robust intelligent control system suitable for complex robotic applications.
在农业等非结构化环境中,由于高度非线性以及对精度和稳定性的要求,控制冗余机械手具有挑战性。本文提出了一种新的智能控制体系结构,该体系结构将基于设计稳定的基于被动的控制(PBC)律与仿生强化学习调谐器相结合。核心新颖之处在于我们的人类学习行为(HLB)模型,与传统的奖励塑造不同,该模型采用有限状态机来动态调节深度确定性策略梯度(DDPG)代理的探索策略。这个状态机在离散状态之间转换,框架为人工情绪(例如,喜悦,恐惧),基于实时性能指标。每个状态直接调整代理的探索噪声,使其能够通过在性能较差时积极探索和在性能较高时利用已知的良好策略来更有效地学习。该自适应调谐器智能地调整PBC定律的比例(Kp)和微分(KD)增益,该定律源自端口控制哈密顿(PCH)模型,以确保潜在的操纵器动力学对于学习代理选择的任何正定增益保持稳定。与固定增益和标准DDPG基线相比,hlb驱动的控制器在模拟的8-DOF机械臂上进行了验证,在轨迹跟踪方面提高了77-80%,能耗降低了30.7%,同时保持了大约10⁻⁴弧度的最终跟踪误差。结果证实,我们基于fsm的自适应探索策略产生了一个高精度、节能、鲁棒的智能控制系统,适用于复杂的机器人应用。
{"title":"Emotion-Based Reinforcement Learning for Passivity-Based Control of Redundant Manipulators in Challenging Environments","authors":"Rami Al-Khulaidi,&nbsp;Tien-Fu Lu,&nbsp;Steven Grainger,&nbsp;Rini Akmeliawati","doi":"10.1016/j.asoc.2026.115997","DOIUrl":"10.1016/j.asoc.2026.115997","url":null,"abstract":"<div><div>Controlling redundant manipulators in unstructured environments like agricultural settings is challenging due to high nonlinearities and the need for both precision and stability. This paper presents a novel intelligent control architecture that integrates a stable-by-design Passivity-Based Control (PBC) law with a bio-inspired reinforcement learning tuner. The core novelty lies in our Human Learning Behaviour (HLB) model, which, unlike conventional reward shaping, employs a finite-state machine to dynamically modulate the exploration strategy of a Deep Deterministic Policy Gradient (DDPG) agent. This state machine transitions between discrete states, framed as artificial emotions (e.g., Joy, Fear), based on real-time performance indicators. Each state directly adjusts the agent's exploration noise, enabling it to learn more efficiently by exploring aggressively when performance is poor and exploiting known good policies when performance is high. This adaptive tuner intelligently adjusts the proportional (K<sub>p</sub>) and differential (K<sub>D</sub>) gains of the PBC law, which is derived from a Port-Controlled Hamiltonian (PCH) model to ensure the underlying manipulator dynamics remain stable for any positive-definite gains selected by the learning agent. Validated on an 8-DOF manipulator in simulated, obstacle-rich agricultural scenarios, the HLB-driven controller demonstrated a 77–80% improvement in trajectory tracking and a 30.7% reduction in energy consumption compared to fixed-gain and standard DDPG baselines, while maintaining a final tracking error of approximately 10⁻⁴ radians. The results confirm that our FSM-based adaptive exploration strategy yields a highly precise, energy-efficient, and robust intelligent control system suitable for complex robotic applications.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"203 ","pages":"Article 115997"},"PeriodicalIF":7.8,"publicationDate":"2026-07-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148575891","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
On directional quantitative assessment of information uncertainty conveyed by interval-valued intuitionistic fuzzy sets 区间值直觉模糊集传递信息不确定性的定向定量评估
IF 7.8 1区 计算机科学 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE Pub Date : 2026-07-17 DOI: 10.1016/j.asoc.2026.115971
Grażyna Szkatuła, Maciej Krawczak
This study presents a novel theoretical and conceptual evaluation framework for interval-valued intuitionistic fuzzy sets consisting of both global and directional components. This enables a more comprehensive analysis of information uncertainty by explicitly considering its structure and variability, treated here as inherent property of fuzzy data. The entire information uncertainty area includes all possible hypothetical changes in the degrees of membership and non-membership under a fixed hesitation parameter. The entire area is divided into three analytically distinct subregions related to changes in (i) the degree of non-membership, (ii) the degree of membership, and (iii) their simultaneous variation. Such decomposition allows for a multi-faceted analysis of the dynamics of information uncertainty, reflecting natural changes perceived from a human perspective. Based on this geometric interpretation, a global quantitative function for assessing information uncertainty and a family of directional assessment functions is developed. These functions assume values within a normalized range and serve as non-probabilistic indicators of both overall and directional information uncertainty. We also propose a ranking procedure based on information certainty. The developed methodology offers a conceptually transparent tool for quantifying information uncertainty, but its broad practical validity still requires confirmation and extensive computational experiments.
本文提出了一种新的由全局分量和方向分量组成的区间值直觉模糊集的理论和概念评价框架。通过明确地考虑其结构和可变性,这使得对信息不确定性的更全面的分析成为可能,这里将其视为模糊数据的固有属性。整个信息不确定性区域包含了在一定犹豫参数下,隶属度和非隶属度的所有可能的假设变化。根据(i)非隶属度的变化,(ii)隶属度的变化,以及(iii)它们的同时变化,整个区域被划分为三个分析上不同的子区域。这种分解允许对信息不确定性的动态进行多方面的分析,反映从人类角度感知的自然变化。在此基础上,建立了信息不确定性评估的全局定量函数和方向性评估函数族。这些函数假设值在一个归一化范围内,并作为总体和方向信息不确定性的非概率指标。我们还提出了一种基于信息确定性的排序方法。所开发的方法为量化信息不确定性提供了概念上透明的工具,但其广泛的实际有效性仍需要确认和大量的计算实验。
{"title":"On directional quantitative assessment of information uncertainty conveyed by interval-valued intuitionistic fuzzy sets","authors":"Grażyna Szkatuła,&nbsp;Maciej Krawczak","doi":"10.1016/j.asoc.2026.115971","DOIUrl":"10.1016/j.asoc.2026.115971","url":null,"abstract":"<div><div>This study presents a novel theoretical and conceptual evaluation framework for interval-valued intuitionistic fuzzy sets consisting of both global and directional components. This enables a more comprehensive analysis of information uncertainty by explicitly considering its structure and variability, treated here as inherent property of fuzzy data. The entire information uncertainty area includes all possible hypothetical changes in the degrees of membership and non-membership under a fixed hesitation parameter. The entire area is divided into three analytically distinct subregions related to changes in (i) the degree of non-membership, (ii) the degree of membership, and (iii) their simultaneous variation. Such decomposition allows for a multi-faceted analysis of the dynamics of information uncertainty, reflecting natural changes perceived from a human perspective. Based on this geometric interpretation, a global quantitative function for assessing information uncertainty and a family of directional assessment functions is developed. These functions assume values within a normalized range and serve as non-probabilistic indicators of both overall and directional information uncertainty. We also propose a ranking procedure based on information certainty. The developed methodology offers a conceptually transparent tool for quantifying information uncertainty, but its broad practical validity still requires confirmation and extensive computational experiments.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"203 ","pages":"Article 115971"},"PeriodicalIF":7.8,"publicationDate":"2026-07-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148575843","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
Measuring successful, failed and deceptive exploration of evolutionary algorithms based on attraction basins 衡量基于吸引力盆地的进化算法探索的成功、失败和欺骗性
IF 7.8 1区 计算机科学 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE Pub Date : 2026-07-15 DOI: 10.1016/j.asoc.2026.115973
Miha Ravber, Jernej Jerebic, Matej Črepinšek, Luka Mernik, Stephen Chen, Marjan Mernik
This paper introduces a novel direct measure of exploration and exploitation in Evolutionary Algorithms, based on attraction basins — regions of the search space with an attractor toward which neighboring solutions tend to evolve. Each solution can be associated with a particular attraction basin. If a new solution belongs to the same attraction basin as its parent, then the search process is identified as exploitation; otherwise, it is identified as exploration. Four different exploration types definitions were used: successful, failed, deceptive, and successful rejection. These are measured directly at the individual level (without precomputing basins) on six single-objective continuous benchmark problems and a real-world problem using DE/rand/1/bin, JADE, LSHADE, ABC, and TLBO. The results revealed that EAs differentiate themselves in how they handle different types of exploration. The approach outperformed prior basin-based measures in efficiency, and exposed the limitations of diversity-based methods, enabling explicit quantification and control of the balance between exploration and exploitation.
本文介绍了进化算法中一种新的直接勘探和开发度量方法,该方法基于吸引盆地——搜索空间中具有吸引子的区域,相邻解倾向于向该区域演化。每个解决方案可以与一个特定的吸引力盆地相关联。如果一个新解与其父解属于同一个吸引盆地,则将搜索过程确定为开发;否则,它就被认为是探索。使用了四种不同的勘探类型定义:成功、失败、欺骗和成功拒绝。使用DE/rand/1/bin、JADE、LSHADE、ABC和TLBO,直接在个人层面(没有预先计算盆地)对六个单目标连续基准问题和一个实际问题进行了测量。结果显示,ea在处理不同类型的探索时表现出了差异性。该方法在效率方面优于先前基于盆地的方法,并暴露了基于多样性的方法的局限性,能够明确量化和控制勘探与开发之间的平衡。
{"title":"Measuring successful, failed and deceptive exploration of evolutionary algorithms based on attraction basins","authors":"Miha Ravber,&nbsp;Jernej Jerebic,&nbsp;Matej Črepinšek,&nbsp;Luka Mernik,&nbsp;Stephen Chen,&nbsp;Marjan Mernik","doi":"10.1016/j.asoc.2026.115973","DOIUrl":"10.1016/j.asoc.2026.115973","url":null,"abstract":"<div><div>This paper introduces a novel direct measure of exploration and exploitation in Evolutionary Algorithms, based on attraction basins — regions of the search space with an attractor toward which neighboring solutions tend to evolve. Each solution can be associated with a particular attraction basin. If a new solution belongs to the same attraction basin as its parent, then the search process is identified as exploitation; otherwise, it is identified as exploration. Four different exploration types definitions were used: successful, failed, deceptive, and successful rejection. These are measured directly at the individual level (without precomputing basins) on six single-objective continuous benchmark problems and a real-world problem using DE/rand/1/bin, JADE, LSHADE, ABC, and TLBO. The results revealed that EAs differentiate themselves in how they handle different types of exploration. The approach outperformed prior basin-based measures in efficiency, and exposed the limitations of diversity-based methods, enabling explicit quantification and control of the balance between exploration and exploitation.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"203 ","pages":"Article 115973"},"PeriodicalIF":7.8,"publicationDate":"2026-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148575918","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
Industrial large model-driven online early fault warning: A novel incremental anomaly detection framework with wavelet scattering contrastive transformer 工业大型模型驱动在线早期故障预警:一种基于小波散射对比变压器的增量异常检测框架
IF 7.8 1区 计算机科学 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE Pub Date : 2026-07-15 DOI: 10.1016/j.asoc.2026.115975
Ziyi Zhang, Yanna Zhang, Wentao Mao, Hui Ma, Jiayi Wang, Shubin Du
Despite substantial potential of industrial large model (ILM), there are still some challenges for deploying ILM in specific industry scenarios: (1) Current ILMs generally adopt learning mode with large-scale parameters, which tends to raise over-fitting on limited data in real-world engineering and fails to generalize to various working conditions; (2) Current ILMs usually employ static deployment based on fine-tuning technique, which is hard to establish a dynamic downstream task with sequentially-collected monitoring data. To address these challenges, this paper selects the early fault detection (EFD) of rotating machinery as an example, and designs an ILM-driven incremental anomaly detection framework that includes a pre-training model and a downstream task. In the pre-training stage, a new wavelet scattering contrastive Transformer (WaveCT) is proposed by introducing wavelet scattering transform to expand physically-meaningful time-frequency feature set from limited data. Through exploring the contextual dependency among normal state data, WaveCT effectively extracts the representation of common one-class detection rules under complex working conditions such as being interference by irregular noise. In the downstream task, an online EFD network is designed based on fine-tuning the sequential data, while an incremental fault alarm indicator is built based on support vector data description (SVDD) generalization error upper bound. Then the statistical distribution shift caused by weak state changes is perceived to achieve early fault warning. Experiments on the vibration signals from our bearing cage testbed and the actual finishing mill bearing of steel enterprise from China validate the effectiveness of the proposed framework.
尽管工业大模型(ILM)具有巨大的潜力,但在具体的工业场景中部署工业大模型仍然存在一些挑战:(1)目前的工业大模型通常采用具有大参数的学习模式,在实际工程中容易引起对有限数据的过拟合,不能推广到各种工况;(2)目前的ilm通常采用基于微调技术的静态部署,难以建立具有顺序采集监测数据的动态下游任务。为了解决这些问题,本文以旋转机械的早期故障检测(EFD)为例,设计了包含预训练模型和下游任务的ilm驱动的增量异常检测框架。在预训练阶段,通过引入小波散射变换,从有限的数据中扩展具有物理意义的时频特征集,提出了一种新的小波散射对比变压器(WaveCT)。通过探索正常状态数据之间的上下文依赖关系,WaveCT有效地提取了受不规则噪声干扰等复杂工况下常见的一类检测规则的表示。在下游任务中,设计了基于序列数据微调的在线EFD网络,并基于支持向量数据描述(SVDD)泛化误差上界构建了增量故障报警指标。然后感知由弱状态变化引起的统计分布偏移,实现早期故障预警。通过对我国轴承保持架试验台和国内某钢铁企业精轧机轴承振动信号的实验,验证了所提框架的有效性。
{"title":"Industrial large model-driven online early fault warning: A novel incremental anomaly detection framework with wavelet scattering contrastive transformer","authors":"Ziyi Zhang,&nbsp;Yanna Zhang,&nbsp;Wentao Mao,&nbsp;Hui Ma,&nbsp;Jiayi Wang,&nbsp;Shubin Du","doi":"10.1016/j.asoc.2026.115975","DOIUrl":"10.1016/j.asoc.2026.115975","url":null,"abstract":"<div><div>Despite substantial potential of industrial large model (ILM), there are still some challenges for deploying ILM in specific industry scenarios: (1) Current ILMs generally adopt learning mode with large-scale parameters, which tends to raise over-fitting on limited data in real-world engineering and fails to generalize to various working conditions; (2) Current ILMs usually employ static deployment based on fine-tuning technique, which is hard to establish a dynamic downstream task with sequentially-collected monitoring data. To address these challenges, this paper selects the early fault detection (EFD) of rotating machinery as an example, and designs an ILM-driven incremental anomaly detection framework that includes a pre-training model and a downstream task. In the pre-training stage, a new wavelet scattering contrastive Transformer (WaveCT) is proposed by introducing wavelet scattering transform to expand physically-meaningful time-frequency feature set from limited data. Through exploring the contextual dependency among normal state data, WaveCT effectively extracts the representation of common one-class detection rules under complex working conditions such as being interference by irregular noise. In the downstream task, an online EFD network is designed based on fine-tuning the sequential data, while an incremental fault alarm indicator is built based on support vector data description (SVDD) generalization error upper bound. Then the statistical distribution shift caused by weak state changes is perceived to achieve early fault warning. Experiments on the vibration signals from our bearing cage testbed and the actual finishing mill bearing of steel enterprise from China validate the effectiveness of the proposed framework.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"203 ","pages":"Article 115975"},"PeriodicalIF":7.8,"publicationDate":"2026-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148582986","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
Distribution and expression aware retrospective learning for single-cell long-tailed class-incremental annotation 单细胞长尾类增量标注的分布和表达感知回溯学习
IF 6.6 1区 计算机科学 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE Pub Date : 2026-05-01 Epub Date: 2026-02-05 DOI: 10.1016/j.asoc.2026.114745
Tianhao Li , Zixuan Wang , Chenpeng Wu , Zhengxiao Huang , Yixin Xiang , Yuhang Liu , Quan Zou , Naifeng Wen , Yongqing Zhang
Single-cell annotation, which identifies specific cell types in biological tissues, is a cornerstone of targeted therapy in precision medicine. However, the lack of incremental annotation capabilities in existing methods limits their adaptability to dynamic immune environment changes and broader applications. Additionally, the high-dimensional sparsity and long-tail distribution of single-cell data hinder practical long-tail incremental annotation. This paper introduces a single-cell incremental annotation framework that combines a distribution-aware diffusion model with an expression-aware knowledge distillation architecture to address these challenges. The distribution-aware module retrospects and replays historical data distributions, enabling precise generation of previously annotated cell types and mitigating catastrophic forgetting. The expression-aware module aligns gene representations through a multi-perspective attention mechanism, enhancing sensitivity to novel cell types while retaining knowledge of dominant categories. A fuzzy incremental guidance mechanism with uncertainty constraints further reduces the adverse effects of long-tail distributions, ensuring more robust annotation performance. Experimental results validate the effectiveness of this framework, demonstrating significant improvements in multi-session incremental annotation accuracy. In addition, this framework enhances the biological and clinical relevance of incremental annotation, enabling continuous integration of new datasets for improved discovery of rare cell types and disease progression analysis. This approach offers a scalable and generalizable solution to the challenges of long-tail incremental annotation in single-cell data.
单细胞注释识别生物组织中特定的细胞类型,是精准医学靶向治疗的基石。然而,现有方法缺乏增量注释能力,限制了其对免疫环境动态变化的适应性和更广泛的应用。此外,单细胞数据的高维稀疏性和长尾分布阻碍了长尾增量标注的实现。本文介绍了一个单细胞增量注释框架,该框架结合了分布感知的扩散模型和表达感知的知识蒸馏体系结构来解决这些挑战。分布感知模块可以回顾和重播历史数据分布,从而精确生成先前注释的细胞类型,并减轻灾难性的遗忘。表达感知模块通过多角度注意机制对齐基因表征,增强对新细胞类型的敏感性,同时保留优势类别的知识。带有不确定性约束的模糊增量引导机制进一步减少了长尾分布的不利影响,确保了更稳健的标注性能。实验结果验证了该框架的有效性,表明该框架在多会话增量标注精度方面有显著提高。此外,该框架增强了增量注释的生物学和临床相关性,使新数据集能够持续集成,以改进罕见细胞类型的发现和疾病进展分析。这种方法为单细胞数据中的长尾增量注释挑战提供了一种可扩展和可推广的解决方案。
{"title":"Distribution and expression aware retrospective learning for single-cell long-tailed class-incremental annotation","authors":"Tianhao Li ,&nbsp;Zixuan Wang ,&nbsp;Chenpeng Wu ,&nbsp;Zhengxiao Huang ,&nbsp;Yixin Xiang ,&nbsp;Yuhang Liu ,&nbsp;Quan Zou ,&nbsp;Naifeng Wen ,&nbsp;Yongqing Zhang","doi":"10.1016/j.asoc.2026.114745","DOIUrl":"10.1016/j.asoc.2026.114745","url":null,"abstract":"<div><div>Single-cell annotation, which identifies specific cell types in biological tissues, is a cornerstone of targeted therapy in precision medicine. However, the lack of incremental annotation capabilities in existing methods limits their adaptability to dynamic immune environment changes and broader applications. Additionally, the high-dimensional sparsity and long-tail distribution of single-cell data hinder practical long-tail incremental annotation. This paper introduces a single-cell incremental annotation framework that combines a distribution-aware diffusion model with an expression-aware knowledge distillation architecture to address these challenges. The distribution-aware module retrospects and replays historical data distributions, enabling precise generation of previously annotated cell types and mitigating catastrophic forgetting. The expression-aware module aligns gene representations through a multi-perspective attention mechanism, enhancing sensitivity to novel cell types while retaining knowledge of dominant categories. A fuzzy incremental guidance mechanism with uncertainty constraints further reduces the adverse effects of long-tail distributions, ensuring more robust annotation performance. Experimental results validate the effectiveness of this framework, demonstrating significant improvements in multi-session incremental annotation accuracy. In addition, this framework enhances the biological and clinical relevance of incremental annotation, enabling continuous integration of new datasets for improved discovery of rare cell types and disease progression analysis. This approach offers a scalable and generalizable solution to the challenges of long-tail incremental annotation in single-cell data.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"193 ","pages":"Article 114745"},"PeriodicalIF":6.6,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146196908","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
Knowledge-guided course recommendation in MOOCs with cross-graph and forgetting-aware cross-attention 基于交叉图和遗忘意识交叉注意的mooc课程推荐
IF 6.6 1区 计算机科学 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE Pub Date : 2026-05-01 Epub Date: 2026-02-03 DOI: 10.1016/j.asoc.2026.114768
Yangtao Zhou , Qingshan Li , Ruoyu Li , Hua Chu , Jianan Li , Feifei Zhu , Xiangming Li , Chengyu Feng , Wanqiang Yang
Course recommendation (CR) plays a pivotal role in the e-learning field, aiming at tailoring course suggestions to learners based on their historical interaction behaviors. Existing CR methods predominantly utilize generic recommendation techniques to infer learners’ preferences from historical interactions. In course learning scenarios, learners’ interaction behaviors are not only preference-driven but also significantly influenced by their knowledge states. However, how to effectively capture the learners’ knowledge states in the CR task remains unexplored, and presents significant challenges due to two critical reasons. First, the knowledge states are intertwined with learners’ preferences, exhibiting mutual collaborative influences. Second, the knowledge states of learners dynamically decay over time due to natural forgetting processes. In this paper, we propose a novel knowledge-guided course recommendation method, named KnowCR, to capture the dynamic knowledge states of learners for enhancing course recommendations. Specifically, we propose a cross-graph collaborative encoding module that integrates the representations of learners’ preferences and knowledge states to exploit their collaborative influences. This module constructs two distinct graphs based on learner-course and learner-exercise interactions, and facilitates information integration via a cross-graph propagation mechanism. Moreover, we devise a forgetting-aware cross-attention encoding module that simulates the decay of short-term and long-term memories to capture the knowledge forgetting properties of learners. This module extracts the dynamic representations of learners by performing sequential encoding with a hierarchical knowledge forgetting mechanism on time-aligned interaction sequences. Extensive experiments on a real-world MOOC dataset demonstrate that our method significantly outperforms the state-of-the-art methods. The implementation codes and datasets are available at https://github.com/KasISET/KnowCR.
课程推荐(CR)在网络学习领域起着举足轻重的作用,其目的是根据学习者的历史互动行为,为学习者量身定制课程建议。现有的CR方法主要利用通用推荐技术从历史交互中推断学习者的偏好。在课程学习情境中,学习者的互动行为不仅受偏好驱动,还受到其知识状态的显著影响。然而,如何有效地捕捉学习者在CR任务中的知识状态仍未得到探索,并且由于两个关键原因而面临重大挑战。首先,知识状态与学习者的偏好交织在一起,表现出相互协作的影响。其次,由于自然遗忘过程,学习者的知识状态会随着时间的推移而动态衰减。在本文中,我们提出了一种新的知识引导课程推荐方法——KnowCR,以捕获学习者的动态知识状态来增强课程推荐。具体而言,我们提出了一个跨图协作编码模块,该模块集成了学习者偏好和知识状态的表示,以利用它们的协作影响。该模块基于学习者-课程和学习者-练习的交互构建两个不同的图,并通过交叉图传播机制促进信息集成。此外,我们设计了一个遗忘意识交叉注意编码模块,模拟短期和长期记忆的衰减,以捕捉学习者的知识遗忘特性。该模块通过在时间对齐的交互序列上使用分层知识遗忘机制进行顺序编码来提取学习者的动态表示。在真实的MOOC数据集上进行的大量实验表明,我们的方法明显优于最先进的方法。实现代码和数据集可在https://github.com/KasISET/KnowCR上获得。
{"title":"Knowledge-guided course recommendation in MOOCs with cross-graph and forgetting-aware cross-attention","authors":"Yangtao Zhou ,&nbsp;Qingshan Li ,&nbsp;Ruoyu Li ,&nbsp;Hua Chu ,&nbsp;Jianan Li ,&nbsp;Feifei Zhu ,&nbsp;Xiangming Li ,&nbsp;Chengyu Feng ,&nbsp;Wanqiang Yang","doi":"10.1016/j.asoc.2026.114768","DOIUrl":"10.1016/j.asoc.2026.114768","url":null,"abstract":"<div><div>Course recommendation (CR) plays a pivotal role in the e-learning field, aiming at tailoring course suggestions to learners based on their historical interaction behaviors. Existing CR methods predominantly utilize generic recommendation techniques to infer learners’ preferences from historical interactions. In course learning scenarios, learners’ interaction behaviors are not only preference-driven but also significantly influenced by their knowledge states. However, how to effectively capture the learners’ knowledge states in the CR task remains unexplored, and presents significant challenges due to two critical reasons. First, the knowledge states are intertwined with learners’ preferences, exhibiting mutual collaborative influences. Second, the knowledge states of learners dynamically decay over time due to natural forgetting processes. In this paper, we propose a novel knowledge-guided course recommendation method, named KnowCR, to capture the dynamic knowledge states of learners for enhancing course recommendations. Specifically, we propose a cross-graph collaborative encoding module that integrates the representations of learners’ preferences and knowledge states to exploit their collaborative influences. This module constructs two distinct graphs based on learner-course and learner-exercise interactions, and facilitates information integration via a cross-graph propagation mechanism. Moreover, we devise a forgetting-aware cross-attention encoding module that simulates the decay of short-term and long-term memories to capture the knowledge forgetting properties of learners. This module extracts the dynamic representations of learners by performing sequential encoding with a hierarchical knowledge forgetting mechanism on time-aligned interaction sequences. Extensive experiments on a real-world MOOC dataset demonstrate that our method significantly outperforms the state-of-the-art methods. The implementation codes and datasets are available at <span><span>https://github.com/KasISET/KnowCR</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"193 ","pages":"Article 114768"},"PeriodicalIF":6.6,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146196904","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
Towards explainable image composition assessment: a dataset and a model 迈向可解释的图像成分评估:一个数据集和一个模型
IF 6.6 1区 计算机科学 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE Pub Date : 2026-05-01 Epub Date: 2026-02-05 DOI: 10.1016/j.asoc.2026.114784
Quan Yuan , Yipo Huang , Pengfei Chen , Leida Li
Explainable Image Composition Assessment (Explainable ICA) provides quantitative scores and qualitative explanations. While recent research on ICA has demonstrated remarkable performance in regressing a scalar score for the overall composition perception, we still need a better understanding of how the image compositions, especially their influencing factors, relate to high-level semantics. However, the requirement for a high-quality explainable ICA dataset limits the broad real-world applications like intelligent photography and content recommendation. This deficiency is primarily because the manual annotation is costly, time-consuming, and struggles to capture the inherent aesthetic subjectivity of composition which heavily relies on expert knowledge. In this work, we take a step toward addressing this challenge through the use of multimodal large language models, and construct a large-scale composition description dataset based on the well-designed prompts in a Chain-of-Thought pattern, named CADB Comments. During its development, we ensured diversity across scenes, composition elements, and levels. This dataset comprises 9497 structured textual annotations. Leveraging this dataset, we propose Multimodal Image Composition Assessment (MICA) to bridge quantitative scoring and qualitative explanation, enabling simultaneous output of scores and descriptions. MICA generates descriptions, embeds composition knowledge into visual encoder, and extracts composition representation using the knowledge-enhanced encoder. A composition level contrastive alignment module strengthens correlation between representation and scores via contrastive learning. Extensive experiments on three datasets demonstrate MICA’s competitive performance in explanation, scoring, and classification. The dataset and code are available at https://github.com/dylanqyuan/MICA.
可解释图像成分评估(可解释ICA)提供定量分数和定性解释。虽然最近的研究表明,ICA在回归整体构图感知的标量分数方面表现出色,但我们仍然需要更好地理解图像构图,特别是其影响因素与高级语义的关系。然而,对高质量可解释的ICA数据集的需求限制了智能摄影和内容推荐等广泛的现实应用。这种缺陷主要是因为手工注释是昂贵的,耗时的,并且很难捕捉到内在的审美主观性,这在很大程度上依赖于专业知识。在这项工作中,我们通过使用多模态大型语言模型向解决这一挑战迈出了一步,并基于思想链模式中精心设计的提示构建了一个大规模的组合描述数据集,名为CADB Comments。在开发过程中,我们确保了场景、构图元素和关卡的多样性。该数据集包括9497个结构化文本注释。利用该数据集,我们提出了多模态图像组成评估(MICA),以架起定量评分和定性解释的桥梁,实现分数和描述的同时输出。MICA生成描述,将组合知识嵌入到视觉编码器中,并使用知识增强编码器提取组合表示。作文水平对比对齐模块通过对比学习增强表征与得分之间的相关性。在三个数据集上的广泛实验证明了MICA在解释、评分和分类方面的竞争力。数据集和代码可在https://github.com/dylanqyuan/MICA上获得。
{"title":"Towards explainable image composition assessment: a dataset and a model","authors":"Quan Yuan ,&nbsp;Yipo Huang ,&nbsp;Pengfei Chen ,&nbsp;Leida Li","doi":"10.1016/j.asoc.2026.114784","DOIUrl":"10.1016/j.asoc.2026.114784","url":null,"abstract":"<div><div>Explainable Image Composition Assessment (Explainable ICA) provides quantitative scores and qualitative explanations. While recent research on ICA has demonstrated remarkable performance in regressing a scalar score for the overall composition perception, we still need a better understanding of how the image compositions, especially their influencing factors, relate to high-level semantics. However, the requirement for a high-quality explainable ICA dataset limits the broad real-world applications like intelligent photography and content recommendation. This deficiency is primarily because the manual annotation is costly, time-consuming, and struggles to capture the inherent aesthetic subjectivity of composition which heavily relies on expert knowledge. In this work, we take a step toward addressing this challenge through the use of multimodal large language models, and construct a large-scale composition description dataset based on the well-designed prompts in a Chain-of-Thought pattern, named CADB Comments. During its development, we ensured diversity across scenes, composition elements, and levels. This dataset comprises 9497 structured textual annotations. Leveraging this dataset, we propose Multimodal Image Composition Assessment (MICA) to bridge quantitative scoring and qualitative explanation, enabling simultaneous output of scores and descriptions. MICA generates descriptions, embeds composition knowledge into visual encoder, and extracts composition representation using the knowledge-enhanced encoder. A composition level contrastive alignment module strengthens correlation between representation and scores via contrastive learning. Extensive experiments on three datasets demonstrate MICA’s competitive performance in explanation, scoring, and classification. The dataset and code are available at <span><span>https://github.com/dylanqyuan/MICA</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"193 ","pages":"Article 114784"},"PeriodicalIF":6.6,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146196902","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
A key node detection method assisted by multi-layer surrogate model and multi-dimensional network features in complex networks 一种基于多层代理模型和多维网络特征的复杂网络关键节点检测方法
IF 6.6 1区 计算机科学 Q1 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE Pub Date : 2026-05-01 Epub Date: 2026-02-10 DOI: 10.1016/j.asoc.2026.114803
Hong Zhao , Ling Tang , Jia-Rui Li , Zong-Gan Chen
Identifying key nodes in complex networks is a critical task, as these nodes significantly shape network structure and functionality. Despite the progress of existing methods, two major challenges remain: high computational complexity in large-scale networks and the insufficient adaptability of evaluation models to diverse network structures. To address these challenges, we propose a key node detection method assisted by multi-layer surrogate model and multi-dimensional network features (KND-MSMNF). Our method introduces a multi-layer surrogate model-assisted network performance evaluation (MSNE) strategy, which integrates a Kriging model and the K-nearest neighbors (KNN) model within a differential evolutionary (DE) algorithm. By dynamically employing different surrogate models at various stages of population evolution, MSNE approximates the objective function, thereby circumventing computationally expensive evaluations and significantly improving the efficiency of network assessment and optimization. Furthermore, an adaptive multi-dimensional network feature search (AMNFS) strategy is proposed to evaluate node importance. This AMNFS strategy dynamically adjusts attribute weights and metrics based on network characteristics, leveraging a comprehensive set of features to avoid bias of single evaluation criteria and enhance adaptability. Additionally, a multimodal optimization-driven key node selection (MKNS) strategy assesses node impact from a global perspective by considering both network cost and overall connectivity. By identifying multiple global optima, MKNS improves the robustness of the solution set and provides flexible alternatives for decision-makers. Extensive experiments on diverse network datasets demonstrate that KND-MSMNF significantly outperforms existing methods in terms of key node detection accuracy (i.e., exceeds 96 %) while maintaining strong practicality and adaptability. Moreover, our method can simultaneously identify multiple combinations of key nodes. This enables the rapid localization of corresponding alternative key nodes in the event of a key node failure, thereby effectively mitigating the risk of network functional impairment caused by the failure of a single node.
识别复杂网络中的关键节点是一项关键任务,因为这些节点显著地决定了网络的结构和功能。尽管现有方法取得了进展,但仍然存在两个主要挑战:大规模网络的计算复杂度高,以及评估模型对不同网络结构的适应性不足。为了解决这些问题,我们提出了一种基于多层代理模型和多维网络特征的关键节点检测方法(KND-MSMNF)。我们的方法引入了一种多层代理模型辅助网络性能评估(MSNE)策略,该策略在差分进化(DE)算法中集成了Kriging模型和k近邻(KNN)模型。MSNE通过在种群进化的不同阶段动态地使用不同的代理模型,逼近目标函数,从而避免了计算量大的评估,显著提高了网络评估和优化的效率。在此基础上,提出了一种自适应多维网络特征搜索(AMNFS)策略来评估节点重要性。该AMNFS策略基于网络特征动态调整属性权重和指标,利用综合的特征集避免单一评价标准的偏差,增强适应性。此外,多模式优化驱动的关键节点选择(MKNS)策略通过考虑网络成本和整体连通性,从全局角度评估节点影响。通过识别多个全局最优,MKNS提高了解决方案集的鲁棒性,为决策者提供了灵活的备选方案。在不同网络数据集上的大量实验表明,KND-MSMNF在保持较强实用性和适应性的同时,在关键节点检测准确率上显著优于现有方法(即超过96% %)。此外,我们的方法可以同时识别多个关键节点的组合。这样可以在关键节点发生故障时快速定位对应的备选关键节点,从而有效降低单个节点故障导致网络功能受损的风险。
{"title":"A key node detection method assisted by multi-layer surrogate model and multi-dimensional network features in complex networks","authors":"Hong Zhao ,&nbsp;Ling Tang ,&nbsp;Jia-Rui Li ,&nbsp;Zong-Gan Chen","doi":"10.1016/j.asoc.2026.114803","DOIUrl":"10.1016/j.asoc.2026.114803","url":null,"abstract":"<div><div>Identifying key nodes in complex networks is a critical task, as these nodes significantly shape network structure and functionality. Despite the progress of existing methods, two major challenges remain: high computational complexity in large-scale networks and the insufficient adaptability of evaluation models to diverse network structures. To address these challenges, we propose a key node detection method assisted by multi-layer surrogate model and multi-dimensional network features (KND-MSMNF). Our method introduces a multi-layer surrogate model-assisted network performance evaluation (MSNE) strategy, which integrates a Kriging model and the <em>K</em>-nearest neighbors (KNN) model within a differential evolutionary (DE) algorithm. By dynamically employing different surrogate models at various stages of population evolution, MSNE approximates the objective function, thereby circumventing computationally expensive evaluations and significantly improving the efficiency of network assessment and optimization. Furthermore, an adaptive multi-dimensional network feature search (AMNFS) strategy is proposed to evaluate node importance. This AMNFS strategy dynamically adjusts attribute weights and metrics based on network characteristics, leveraging a comprehensive set of features to avoid bias of single evaluation criteria and enhance adaptability. Additionally, a multimodal optimization-driven key node selection (MKNS) strategy assesses node impact from a global perspective by considering both network cost and overall connectivity. By identifying multiple global optima, MKNS improves the robustness of the solution set and provides flexible alternatives for decision-makers. Extensive experiments on diverse network datasets demonstrate that KND-MSMNF significantly outperforms existing methods in terms of key node detection accuracy (i.e., exceeds 96 %) while maintaining strong practicality and adaptability. Moreover, our method can simultaneously identify multiple combinations of key nodes. This enables the rapid localization of corresponding alternative key nodes in the event of a key node failure, thereby effectively mitigating the risk of network functional impairment caused by the failure of a single node.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"193 ","pages":"Article 114803"},"PeriodicalIF":6.6,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146196905","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
期刊
Applied Soft Computing
全部 Acc. Chem. Res. ACS Applied Bio Materials ACS Appl. Electron. Mater. ACS Appl. Energy Mater. ACS Appl. Mater. Interfaces ACS Appl. Nano Mater. ACS Appl. Polym. Mater. ACS BIOMATER-SCI ENG ACS Catal. ACS Cent. Sci. ACS Chem. Biol. ACS Chemical Health & Safety ACS Chem. Neurosci. ACS Comb. Sci. ACS Earth Space Chem. ACS Energy Lett. ACS Infect. Dis. ACS Macro Lett. ACS Mater. Lett. ACS Med. Chem. Lett. ACS Nano ACS Omega ACS Photonics ACS Sens. ACS Sustainable Chem. Eng. ACS Synth. Biol. Anal. Chem. BIOCHEMISTRY-US Bioconjugate Chem. BIOMACROMOLECULES Chem. Res. Toxicol. Chem. Rev. Chem. Mater. CRYST GROWTH DES ENERG FUEL Environ. Sci. Technol. Environ. Sci. Technol. Lett. Eur. J. Inorg. Chem. IND ENG CHEM RES Inorg. Chem. J. Agric. Food. Chem. J. Chem. Eng. Data J. Chem. Educ. J. Chem. Inf. Model. J. Chem. Theory Comput. J. Med. Chem. J. Nat. Prod. J PROTEOME RES J. Am. Chem. Soc. LANGMUIR MACROMOLECULES Mol. Pharmaceutics Nano Lett. Org. Lett. ORG PROCESS RES DEV ORGANOMETALLICS J. Org. Chem. J. Phys. Chem. J. Phys. Chem. A J. Phys. Chem. B J. Phys. Chem. C J. Phys. Chem. Lett. Analyst Anal. Methods Biomater. Sci. Catal. Sci. Technol. Chem. Commun. Chem. Soc. Rev. CHEM EDUC RES PRACT CRYSTENGCOMM Dalton Trans. Energy Environ. Sci. ENVIRON SCI-NANO ENVIRON SCI-PROC IMP ENVIRON SCI-WAT RES Faraday Discuss. Food Funct. Green Chem. Inorg. Chem. Front. Integr. Biol. J. Anal. At. Spectrom. J. Mater. Chem. A J. Mater. Chem. B J. Mater. Chem. C Lab Chip Mater. Chem. Front. Mater. Horiz. MEDCHEMCOMM Metallomics Mol. Biosyst. Mol. Syst. Des. Eng. Nanoscale Nanoscale Horiz. Nat. Prod. Rep. New J. Chem. Org. Biomol. Chem. Org. Chem. Front. PHOTOCH PHOTOBIO SCI PCCP Polym. Chem.
﹀
mjeditor
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
0
微信
客服QQ
Book学术公众号 扫码关注我们
反馈
×
意见反馈
请填写您的意见或建议
请填写您的手机或邮箱
×
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
现在去查看 取消
×
提示
确定
Book学术官方微信
Book学术官方微信
Book学术文献互助
Book学术文献互助群
群 号:604180095
Book学术
文献互助 智能选刊 最新文献 互助须知 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 Book学术 All rights reserved.
ghs 京公网安备 11010802042870号 京ICP备2023020795号-1