The outstanding growth in the applications of large language models(LLMs)demonstrates the significance of adaptive and efficient prompt engineering tactics.The existing methods may not be variable,vigorous and streaml...The outstanding growth in the applications of large language models(LLMs)demonstrates the significance of adaptive and efficient prompt engineering tactics.The existing methods may not be variable,vigorous and streamlined in different domains.The offered study introduces an immediate optimization outline,named PROMPTx-PE,that is going to yield a greater level of precision and strength when it comes to the assignments that are premised on LLM.The proposed systemfeatures a timely selection schemewhich is informed by reinforcement learning,a contextual layer and a dynamic weighting module which is regulated by Lyapunov-based stability guidelines.The PROMPTx-PE dynamically varies the exploration and exploitation of the prompt space,depending on real-time feedback and multi-objective reward development.Extensive testing on both benchmark(GLUE,SuperGLUE)and domain-specific data(Healthcare-QA and Industrial-NER)demonstrates a large best performance to be 89.4%and a strong robustness disconnect with under 3%computation expense.The results confirm the effectiveness,consistency,and scalability of PROMPTx-PE as a platform of adaptive prompt engineering based on recent uses of LLMs.展开更多
Evaluating the adversarial robustness of classification algorithms in machine learning is a crucial domain.However,current methods lack measurable and interpretable metrics.To address this issue,this paper introduces ...Evaluating the adversarial robustness of classification algorithms in machine learning is a crucial domain.However,current methods lack measurable and interpretable metrics.To address this issue,this paper introduces a visual evaluation index named confidence centroid skewing quadrilateral,which is based on a classification confidence-based confusion matrix,offering a quantitative and visual comparison of the adversarial robustness among different classification algorithms,and enhances intuitiveness and interpretability of attack impacts.We first conduct a validity test and sensitive analysis of the method.Then,prove its effectiveness through the experiments of five classification algorithms including artificial neural network(ANN),logistic regression(LR),support vector machine(SVM),convolutional neural network(CNN)and transformer against three adversarial attacks such as fast gradient sign method(FGSM),DeepFool,and projected gradient descent(PGD)attack.展开更多
Complex network models are frequently employed for simulating and studyingdiverse real-world complex systems.Among these models,scale-free networks typically exhibit greater fragility to malicious attacks.Consequently...Complex network models are frequently employed for simulating and studyingdiverse real-world complex systems.Among these models,scale-free networks typically exhibit greater fragility to malicious attacks.Consequently,enhancing the robustness of scale-free networks has become a pressing issue.To address this problem,this paper proposes a Multi-Granularity Integration Algorithm(MGIA),which aims to improve the robustness of scale-free networks while keeping the initial degree of each node unchanged,ensuring network connectivity and avoiding the generation of multiple edges.The algorithm generates a multi-granularity structure from the initial network to be optimized,then uses different optimization strategies to optimize the networks at various granular layers in this structure,and finally realizes the information exchange between different granular layers,thereby further enhancing the optimization effect.We propose new network refresh,crossover,and mutation operators to ensure that the optimized network satisfies the given constraints.Meanwhile,we propose new network similarity and network dissimilarity evaluation metrics to improve the effectiveness of the optimization operators in the algorithm.In the experiments,the MGIA enhances the robustness of the scale-free network by 67.6%.This improvement is approximately 17.2%higher than the optimization effects achieved by eight currently existing complex network robustness optimization algorithms.展开更多
While in the past the robustness of transportation networks was studied considering the cyber and physical space as isolated environments this is no longer the case.Integrating the Internet of Things devices in the se...While in the past the robustness of transportation networks was studied considering the cyber and physical space as isolated environments this is no longer the case.Integrating the Internet of Things devices in the sensing area of transportation infrastructure has resulted in ubiquitous cyber-physical systems and increasing interdependen-cies between the physical and cyber networks.As a result,the robustness of transportation networks relies on the uninterrupted serviceability of physical and cyber networks.Current studies on interdependent networks overlook the civil engineering aspect of cyber-physical systems.Firstly,they rely on the assumption of a uniform and strong level of interdependency.That is,once a node within a network fails its counterpart fails immedi-ately.Current studies overlook the impact of earthquake and other natural hazards on the operation of modern transportation infrastructure,that now serve as a cyber-physical system.The last is responsible not only for the physical operation(e.g.,flow of vehicles)but also for the continuous data transmission and subsequently the cy-ber operation of the entire transportation network.Therefore,the robustness of modern transportation networks should be modelled from a new cyber-physical perspective that includes civil engineering aspects.In this paper,we propose a new robustness assessment approach for modern transportation networks and their underlying in-terdependent physical and cyber network,subjected to earthquake events.The novelty relies on the modelling of interdependent networks,in the form of a graph,based on their interdependency levels.We associate the service-ability level of the coupled physical and cyber network with the damage states induced by earthquake events.Robustness is then measured as a degradation of the cyber-physical serviceability level.The application of the approach is demonstrated by studying an illustrative transportation network using seismic data from real-world transportation infrastructure.Furthermore,we propose the integration of a robustness improvement indicator based on physical and cyber attributes to enhance the cyber-physical serviceability level.Results indicate an improvement in robustness level(i.e.,41%)by adopting the proposed robustness improvement indicator.The usefulness of our approach is highlighted by comparing it with other methods that consider strong interdepen-dencies and key node protection strategies.The approach is of interest to stakeholders who are attempting to incorporate cyber-physical systems into civil engineering systems.展开更多
Numerous uncertainties in practical production and operation can seriously affect the drive performance of permanent magnet synchronous machines(PMSMs).Various robust control methods have been developed to mitigate or...Numerous uncertainties in practical production and operation can seriously affect the drive performance of permanent magnet synchronous machines(PMSMs).Various robust control methods have been developed to mitigate or eliminate the effects of these uncertainties.However,the robustness to uncertainties of electrical drive systems has not been clearly defined.No systemic procedures have been proposed to evaluate a control system's robustness(how robust it is).This paper proposes a systemic method for evaluating control systems'robustness to uncertainties.The concept and fundamental theory of robust control are illustrated by considering a simple uncertain feedback control system.The effects of uncertainties on the control performance and stability are analyzed and discussed.The concept of design for six-sigma(a robust design method)is employed to numerically evaluate the robustness levels of control systems.To show the effectiveness of the proposed robustness evaluation method,case studies are conducted for second-order systems,DC motor drive systems,and PMSM drive systems.Besides the conventional predictive control of PMSM drive,three different robust predictive control methods are evaluated in terms of two different parametric uncertainty ranges and three application requirements against parametric uncertainties.展开更多
Software systems play increasing important roles in modern society,and the ability against attacks is of great practical importance to crucial software systems,resulting in that the structure and robustness of softwar...Software systems play increasing important roles in modern society,and the ability against attacks is of great practical importance to crucial software systems,resulting in that the structure and robustness of software systems have attracted a tremendous amount of interest in recent years.In this paper,based on the source code of Tar and MySQL,we propose an approach to generate coupled software networks and construct three kinds of directed software networks:The function call network,the weakly coupled network and the strongly coupled network.The structural properties of these complex networks are extensively investigated.It is found that the average influence and the average dependence for all functions are the same.Moreover,eight attacking strategies and two robustness indicators(the weakly connected indicator and the strongly connected indicator)are introduced to analyze the robustness of software networks.This shows that the strongly coupled network is just a weakly connected network rather than a strongly connected one.For MySQL,high in-degree strategy outperforms other attacking strategies when the weakly connected indicator is used.On the other hand,high out-degree strategy is a good choice when the strongly connected indicator is adopted.This work will highlight a better understanding of the structure and robustness of software networks.展开更多
Superhydrophobic glass has inspiring development prospects in endoscopes,solar panels and other engineering and medical fields.However,the surface topography required to achieve superhydrophobicity will inevitably aff...Superhydrophobic glass has inspiring development prospects in endoscopes,solar panels and other engineering and medical fields.However,the surface topography required to achieve superhydrophobicity will inevitably affect the surface transparency and limit the application of glass materials.To resolve the contradiction between the surface transparency and the robust superhydrophobicity,an efficient and low-cost laser-chemical surface functionalization process was utilized to fabricate superhydrophobic glass surface.The results show that the air can be effectively trapped in surface micro/nanostructure induced by laser texturing,thus reducing the solid-liquid contact area and interfacial tension.The deposition of hydrophobic carbon-containing groups on the surface can be accelerated by chemical treatment,and the surface energy is significantly reduced.The glass surface exhibits marvelous robust superhydrophobicity with a contact angle of 155.8°and a roll-off angle of 7.2°under the combination of hierarchical micro/nanostructure and low surface energy.Moreover,the surface transparency of the prepared superhydrophobic glass was only 5.42%lower than that of the untreated surface.This superhydrophobic glass with high transparency still maintains excellent superhydrophobicity after durability and stability tests.The facile fabrication of superhydrophobic glass with high transparency and robustness provides a strong reference for further expanding the application value of glass materials.展开更多
Tag recommendation systems can significantly improve the accuracy of information retrieval by recommending relevant tag sets that align with user preferences and resource characteristics.However,metric learning method...Tag recommendation systems can significantly improve the accuracy of information retrieval by recommending relevant tag sets that align with user preferences and resource characteristics.However,metric learning methods often suffer from high sensitivity,leading to unstable recommendation results when facing adversarial samples generated through malicious user behavior.Adversarial training is considered to be an effective method for improving the robustness of tag recommendation systems and addressing adversarial samples.However,it still faces the challenge of overfitting.Although curriculum learning-based adversarial training somewhat mitigates this issue,challenges still exist,such as the lack of a quantitative standard for attack intensity and catastrophic forgetting.To address these challenges,we propose a Self-Paced Adversarial Metric Learning(SPAML)method.First,we employ a metric learning model to capture the deep distance relationships between normal samples.Then,we incorporate a self-paced adversarial training model,which dynamically adjusts the weights of adversarial samples,allowing the model to progressively learn from simpler to more complex adversarial samples.Finally,we jointly optimize the metric learning loss and self-paced adversarial training loss in an adversarial manner,enhancing the robustness and performance of tag recommendation tasks.Extensive experiments on the MovieLens and LastFm datasets demonstrate that SPAML achieves F1@3 and NDCG@3 scores of 22%and 32.7%on the MovieLens dataset,and 19.4%and 29%on the LastFm dataset,respectively,outperforming the most competitive baselines.Specifically,F1@3 improves by 4.7%and 6.8%,and NDCG@3 improves by 5.0%and 6.9%,respectively.展开更多
Achieving high-resolution intracranial imaging in a safe and portable manner is critical for the diagnosis of intracranial diseases,preoperative planning of craniotomies and intraoperative management during craniotomy...Achieving high-resolution intracranial imaging in a safe and portable manner is critical for the diagnosis of intracranial diseases,preoperative planning of craniotomies and intraoperative management during craniotomy procedures.Adaptive waveform inversion(AWI),a variant of full waveform inversion(FWI),has shown potential in intracranial ultrasound imaging.However,the robustness of AWI is affected by the parameterization of the Gaussian penalty matrix and the challenges posed by transcranial scenarios.Conventional AWI struggles to produce accurate images in these cases,limiting its application in critical medical settings.To address these issues,we propose a stabilized adaptive waveform inversion(SAWI)method,which introduces a user-defined zero-lag position for theWiener filter.Numerical experiments demonstrate that SAWI can achieve accurate imaging under Gaussian penalty matrix parameter settings where AWI fails,perform successful transcranial imaging in configurations where AWI cannot,and maintain the same imaging accuracy as AWI.The advantage of this method is that it achieves these advancements without modifying the AWI framework or increasing computational costs,which helps to promote the application of AWI in medical fields,particularly in transcranial scenarios.展开更多
Time series anomaly detection is critical in domains such as manufacturing,finance,and cybersecurity.Recent generative AI models,particularly Transformer-and Autoencoder-based architectures,show strong accuracy but th...Time series anomaly detection is critical in domains such as manufacturing,finance,and cybersecurity.Recent generative AI models,particularly Transformer-and Autoencoder-based architectures,show strong accuracy but their robustness under noisy conditions is less understood.This study evaluates three representative models—AnomalyTransformer,TranAD,and USAD—on the Server Machine Dataset(SMD)and cross-domain benchmarks including the SoilMoisture Active Passive(SMAP)dataset,theMars Science Laboratory(MSL)dataset,and the Secure Water Treatment(SWaT)testbed.Seven noise settings(five canonical,two mixed)at multiple intensities are tested under fixed clean-data training,with variations in window,stride,and thresholding.Results reveal distinct robustness profiles:AnomalyTransformermaintains recall but loses precision under abrupt noise,TranAD balances sensitivity yet is vulnerable to structured anomalies,and USAD resists Gaussian perturbations but collapses under block anomalies.Quantitatively,F1 drops 60%–70%on noisy SMD,with severe collapse in SWaT(F1≤0.10,Drop up to 84%)but relative stability on SMAP/MSL(Drop within±10%).Overall,generative models exhibit complementary robustness patterns,highlighting noise-type dependent vulnerabilities and providing practical guidance for robust deployment.展开更多
新型电力系统建设背景下,新能源渗透率不断提高造成电力系统灵活性需求剧增。针对电力系统灵活性供需失衡问题,提出含碳捕集电厂的源荷储资源数据驱动鲁棒优化调度模型。首先,基于碳捕集电厂、抽水蓄能等源荷储灵活性资源运行特性,刻画...新型电力系统建设背景下,新能源渗透率不断提高造成电力系统灵活性需求剧增。针对电力系统灵活性供需失衡问题,提出含碳捕集电厂的源荷储资源数据驱动鲁棒优化调度模型。首先,基于碳捕集电厂、抽水蓄能等源荷储灵活性资源运行特性,刻画电力系统灵活性供给能力;其次,采用数据驱动方法构建椭球不确定性集以刻画风电、光伏波动区间,根据各时刻边界值量化灵活性需求;进而,结合灵活性供需关系,提出源荷储资源数据驱动鲁棒优化模型,并采用改进列与约束生成算法(column and constraint generation,C&CG)进行求解。算例仿真表明,通过协同调度源荷储灵活性资源有助于支撑电力系统功率平衡,提高灵活性裕度,数据驱动鲁棒优化方法能够剔除传统盒式不确定集中的不实际恶劣场景,进而改善传统鲁棒优化模型过于保守的问题,并显著提升计算效率。展开更多
基金supported by the National Science and Technology Council(NSTC),Taiwan,under grant number 114-2221-E-182-041-MY3by Chang Gung University and Chang Gung Memorial Hospital under project number NERPD4Q0021.
文摘The outstanding growth in the applications of large language models(LLMs)demonstrates the significance of adaptive and efficient prompt engineering tactics.The existing methods may not be variable,vigorous and streamlined in different domains.The offered study introduces an immediate optimization outline,named PROMPTx-PE,that is going to yield a greater level of precision and strength when it comes to the assignments that are premised on LLM.The proposed systemfeatures a timely selection schemewhich is informed by reinforcement learning,a contextual layer and a dynamic weighting module which is regulated by Lyapunov-based stability guidelines.The PROMPTx-PE dynamically varies the exploration and exploitation of the prompt space,depending on real-time feedback and multi-objective reward development.Extensive testing on both benchmark(GLUE,SuperGLUE)and domain-specific data(Healthcare-QA and Industrial-NER)demonstrates a large best performance to be 89.4%and a strong robustness disconnect with under 3%computation expense.The results confirm the effectiveness,consistency,and scalability of PROMPTx-PE as a platform of adaptive prompt engineering based on recent uses of LLMs.
文摘Evaluating the adversarial robustness of classification algorithms in machine learning is a crucial domain.However,current methods lack measurable and interpretable metrics.To address this issue,this paper introduces a visual evaluation index named confidence centroid skewing quadrilateral,which is based on a classification confidence-based confusion matrix,offering a quantitative and visual comparison of the adversarial robustness among different classification algorithms,and enhances intuitiveness and interpretability of attack impacts.We first conduct a validity test and sensitive analysis of the method.Then,prove its effectiveness through the experiments of five classification algorithms including artificial neural network(ANN),logistic regression(LR),support vector machine(SVM),convolutional neural network(CNN)and transformer against three adversarial attacks such as fast gradient sign method(FGSM),DeepFool,and projected gradient descent(PGD)attack.
基金National Natural Science Foundation of China(11971211,12171388).
文摘Complex network models are frequently employed for simulating and studyingdiverse real-world complex systems.Among these models,scale-free networks typically exhibit greater fragility to malicious attacks.Consequently,enhancing the robustness of scale-free networks has become a pressing issue.To address this problem,this paper proposes a Multi-Granularity Integration Algorithm(MGIA),which aims to improve the robustness of scale-free networks while keeping the initial degree of each node unchanged,ensuring network connectivity and avoiding the generation of multiple edges.The algorithm generates a multi-granularity structure from the initial network to be optimized,then uses different optimization strategies to optimize the networks at various granular layers in this structure,and finally realizes the information exchange between different granular layers,thereby further enhancing the optimization effect.We propose new network refresh,crossover,and mutation operators to ensure that the optimized network satisfies the given constraints.Meanwhile,we propose new network similarity and network dissimilarity evaluation metrics to improve the effectiveness of the optimization operators in the algorithm.In the experiments,the MGIA enhances the robustness of the scale-free network by 67.6%.This improvement is approximately 17.2%higher than the optimization effects achieved by eight currently existing complex network robustness optimization algorithms.
文摘While in the past the robustness of transportation networks was studied considering the cyber and physical space as isolated environments this is no longer the case.Integrating the Internet of Things devices in the sensing area of transportation infrastructure has resulted in ubiquitous cyber-physical systems and increasing interdependen-cies between the physical and cyber networks.As a result,the robustness of transportation networks relies on the uninterrupted serviceability of physical and cyber networks.Current studies on interdependent networks overlook the civil engineering aspect of cyber-physical systems.Firstly,they rely on the assumption of a uniform and strong level of interdependency.That is,once a node within a network fails its counterpart fails immedi-ately.Current studies overlook the impact of earthquake and other natural hazards on the operation of modern transportation infrastructure,that now serve as a cyber-physical system.The last is responsible not only for the physical operation(e.g.,flow of vehicles)but also for the continuous data transmission and subsequently the cy-ber operation of the entire transportation network.Therefore,the robustness of modern transportation networks should be modelled from a new cyber-physical perspective that includes civil engineering aspects.In this paper,we propose a new robustness assessment approach for modern transportation networks and their underlying in-terdependent physical and cyber network,subjected to earthquake events.The novelty relies on the modelling of interdependent networks,in the form of a graph,based on their interdependency levels.We associate the service-ability level of the coupled physical and cyber network with the damage states induced by earthquake events.Robustness is then measured as a degradation of the cyber-physical serviceability level.The application of the approach is demonstrated by studying an illustrative transportation network using seismic data from real-world transportation infrastructure.Furthermore,we propose the integration of a robustness improvement indicator based on physical and cyber attributes to enhance the cyber-physical serviceability level.Results indicate an improvement in robustness level(i.e.,41%)by adopting the proposed robustness improvement indicator.The usefulness of our approach is highlighted by comparing it with other methods that consider strong interdepen-dencies and key node protection strategies.The approach is of interest to stakeholders who are attempting to incorporate cyber-physical systems into civil engineering systems.
文摘Numerous uncertainties in practical production and operation can seriously affect the drive performance of permanent magnet synchronous machines(PMSMs).Various robust control methods have been developed to mitigate or eliminate the effects of these uncertainties.However,the robustness to uncertainties of electrical drive systems has not been clearly defined.No systemic procedures have been proposed to evaluate a control system's robustness(how robust it is).This paper proposes a systemic method for evaluating control systems'robustness to uncertainties.The concept and fundamental theory of robust control are illustrated by considering a simple uncertain feedback control system.The effects of uncertainties on the control performance and stability are analyzed and discussed.The concept of design for six-sigma(a robust design method)is employed to numerically evaluate the robustness levels of control systems.To show the effectiveness of the proposed robustness evaluation method,case studies are conducted for second-order systems,DC motor drive systems,and PMSM drive systems.Besides the conventional predictive control of PMSM drive,three different robust predictive control methods are evaluated in terms of two different parametric uncertainty ranges and three application requirements against parametric uncertainties.
基金supported by the Beijing Education Commission Science and Technology Project(No.KM201811417005)the National Natural Science Foundation of China(No.62173237)+6 种基金the Aeronautical Science Foundation of China(No.20240055054001)the Open Fund of State Key Laboratory of Satellite Navigation System and Equipment Technology(No.CEPNT2023A01)Joint Fund of Ministry of Natural Resources Key Laboratory of Spatiotemporal Perception and Intelligent Processing(No.232203)the Civil Aviation Flight Technology and Flight Safety Engineering Technology Research Center of Sichuan(No.GY2024-02B)the Applied Basic Research Programs of Liaoning Province(No.2025JH2/101300011)the General Project of Liaoning Provincial Education Department(No.20250054)Research on Safety Intelligent Management Technology and Systems for Mixed Operations of General Aviation Aircraft in Low-Altitude Airspace(No.310125011).
文摘Software systems play increasing important roles in modern society,and the ability against attacks is of great practical importance to crucial software systems,resulting in that the structure and robustness of software systems have attracted a tremendous amount of interest in recent years.In this paper,based on the source code of Tar and MySQL,we propose an approach to generate coupled software networks and construct three kinds of directed software networks:The function call network,the weakly coupled network and the strongly coupled network.The structural properties of these complex networks are extensively investigated.It is found that the average influence and the average dependence for all functions are the same.Moreover,eight attacking strategies and two robustness indicators(the weakly connected indicator and the strongly connected indicator)are introduced to analyze the robustness of software networks.This shows that the strongly coupled network is just a weakly connected network rather than a strongly connected one.For MySQL,high in-degree strategy outperforms other attacking strategies when the weakly connected indicator is used.On the other hand,high out-degree strategy is a good choice when the strongly connected indicator is adopted.This work will highlight a better understanding of the structure and robustness of software networks.
基金Projects(52105175,52305149)supported by the National Natural Science Foundation of ChinaProject(2242024RCB0035)supported by the Zhishan Young Scholar Program of Southeast University,China+5 种基金Project(BK20210235)supported by the Natural Science Foundation of Jiangsu Province,ChinaProject(2023MK042)supported by the State Administration for Market Regulation,ChinaProject(KJ2023003)supported by the Jiangsu Administration for Market Regulation,ChinaProjects(KJ(Y)202429,KJ(YJ)2023001)supported by the Jiangsu Province Special Equipment Safety Supervision Inspection Institute,ChinaProject(JSSCBS20210121)supported by the Jiangsu Provincial Innovative and Entrepreneurial Doctor Program,ChinaProject(1102002310)supported by the Technology Innovation Project for Returnees in Nanjing,China。
文摘Superhydrophobic glass has inspiring development prospects in endoscopes,solar panels and other engineering and medical fields.However,the surface topography required to achieve superhydrophobicity will inevitably affect the surface transparency and limit the application of glass materials.To resolve the contradiction between the surface transparency and the robust superhydrophobicity,an efficient and low-cost laser-chemical surface functionalization process was utilized to fabricate superhydrophobic glass surface.The results show that the air can be effectively trapped in surface micro/nanostructure induced by laser texturing,thus reducing the solid-liquid contact area and interfacial tension.The deposition of hydrophobic carbon-containing groups on the surface can be accelerated by chemical treatment,and the surface energy is significantly reduced.The glass surface exhibits marvelous robust superhydrophobicity with a contact angle of 155.8°and a roll-off angle of 7.2°under the combination of hierarchical micro/nanostructure and low surface energy.Moreover,the surface transparency of the prepared superhydrophobic glass was only 5.42%lower than that of the untreated surface.This superhydrophobic glass with high transparency still maintains excellent superhydrophobicity after durability and stability tests.The facile fabrication of superhydrophobic glass with high transparency and robustness provides a strong reference for further expanding the application value of glass materials.
基金supported by the Key Research and Development Program of Zhejiang Province(No.2024C01071)the Natural Science Foundation of Zhejiang Province(No.LQ15F030006).
文摘Tag recommendation systems can significantly improve the accuracy of information retrieval by recommending relevant tag sets that align with user preferences and resource characteristics.However,metric learning methods often suffer from high sensitivity,leading to unstable recommendation results when facing adversarial samples generated through malicious user behavior.Adversarial training is considered to be an effective method for improving the robustness of tag recommendation systems and addressing adversarial samples.However,it still faces the challenge of overfitting.Although curriculum learning-based adversarial training somewhat mitigates this issue,challenges still exist,such as the lack of a quantitative standard for attack intensity and catastrophic forgetting.To address these challenges,we propose a Self-Paced Adversarial Metric Learning(SPAML)method.First,we employ a metric learning model to capture the deep distance relationships between normal samples.Then,we incorporate a self-paced adversarial training model,which dynamically adjusts the weights of adversarial samples,allowing the model to progressively learn from simpler to more complex adversarial samples.Finally,we jointly optimize the metric learning loss and self-paced adversarial training loss in an adversarial manner,enhancing the robustness and performance of tag recommendation tasks.Extensive experiments on the MovieLens and LastFm datasets demonstrate that SPAML achieves F1@3 and NDCG@3 scores of 22%and 32.7%on the MovieLens dataset,and 19.4%and 29%on the LastFm dataset,respectively,outperforming the most competitive baselines.Specifically,F1@3 improves by 4.7%and 6.8%,and NDCG@3 improves by 5.0%and 6.9%,respectively.
基金supported by the National Natural Science Foundation of China(Grant No.82151302)the National High Level Hospital Clinical Research Funding(Grant No.2022-PUMCH-B-113)+1 种基金the National High Level Hospital Clinical Research Funding(Grant No.2022-PUMCH-A-019)the CAMS Innovation Fund for Medical Sciences(Grant No.2021-12M-1-014).
文摘Achieving high-resolution intracranial imaging in a safe and portable manner is critical for the diagnosis of intracranial diseases,preoperative planning of craniotomies and intraoperative management during craniotomy procedures.Adaptive waveform inversion(AWI),a variant of full waveform inversion(FWI),has shown potential in intracranial ultrasound imaging.However,the robustness of AWI is affected by the parameterization of the Gaussian penalty matrix and the challenges posed by transcranial scenarios.Conventional AWI struggles to produce accurate images in these cases,limiting its application in critical medical settings.To address these issues,we propose a stabilized adaptive waveform inversion(SAWI)method,which introduces a user-defined zero-lag position for theWiener filter.Numerical experiments demonstrate that SAWI can achieve accurate imaging under Gaussian penalty matrix parameter settings where AWI fails,perform successful transcranial imaging in configurations where AWI cannot,and maintain the same imaging accuracy as AWI.The advantage of this method is that it achieves these advancements without modifying the AWI framework or increasing computational costs,which helps to promote the application of AWI in medical fields,particularly in transcranial scenarios.
基金supported by the“Regional Innovation System&Education(RISE)”through the Seoul RISE Center,funded by the Ministry of Education(MOE)the Seoul Metropolitan Government(2025-RISE-01-018-04)supported by the Korea Digital Forensic Center.
文摘Time series anomaly detection is critical in domains such as manufacturing,finance,and cybersecurity.Recent generative AI models,particularly Transformer-and Autoencoder-based architectures,show strong accuracy but their robustness under noisy conditions is less understood.This study evaluates three representative models—AnomalyTransformer,TranAD,and USAD—on the Server Machine Dataset(SMD)and cross-domain benchmarks including the SoilMoisture Active Passive(SMAP)dataset,theMars Science Laboratory(MSL)dataset,and the Secure Water Treatment(SWaT)testbed.Seven noise settings(five canonical,two mixed)at multiple intensities are tested under fixed clean-data training,with variations in window,stride,and thresholding.Results reveal distinct robustness profiles:AnomalyTransformermaintains recall but loses precision under abrupt noise,TranAD balances sensitivity yet is vulnerable to structured anomalies,and USAD resists Gaussian perturbations but collapses under block anomalies.Quantitatively,F1 drops 60%–70%on noisy SMD,with severe collapse in SWaT(F1≤0.10,Drop up to 84%)but relative stability on SMAP/MSL(Drop within±10%).Overall,generative models exhibit complementary robustness patterns,highlighting noise-type dependent vulnerabilities and providing practical guidance for robust deployment.
文摘新型电力系统建设背景下,新能源渗透率不断提高造成电力系统灵活性需求剧增。针对电力系统灵活性供需失衡问题,提出含碳捕集电厂的源荷储资源数据驱动鲁棒优化调度模型。首先,基于碳捕集电厂、抽水蓄能等源荷储灵活性资源运行特性,刻画电力系统灵活性供给能力;其次,采用数据驱动方法构建椭球不确定性集以刻画风电、光伏波动区间,根据各时刻边界值量化灵活性需求;进而,结合灵活性供需关系,提出源荷储资源数据驱动鲁棒优化模型,并采用改进列与约束生成算法(column and constraint generation,C&CG)进行求解。算例仿真表明,通过协同调度源荷储灵活性资源有助于支撑电力系统功率平衡,提高灵活性裕度,数据驱动鲁棒优化方法能够剔除传统盒式不确定集中的不实际恶劣场景,进而改善传统鲁棒优化模型过于保守的问题,并显著提升计算效率。
