检索结果-内蒙古大学图书馆

Unknown-Aware Bilateral Dependency Optimization for Defending Against Model Inversion Attacks

IEEE Transactions on Pattern Analysis and Machine Intelligence 2025年 PP卷 PP页

作者： Peng, Xiong Liu, Feng Wang, Nannan Lan, Long Liu, Tongliang Cheung, Yiu-Ming Han, Bo Hong Kong Baptist University Department of Computer Science Hong Kong The University of Melbourne School of Computing and Information Systems Australia Xidian University State Key Laboratory of Integrated Services Networks School of Telecommunications Engineering Shaanxi Xi'an710071 China National University of Defense Technology State Key Laboratory of High Performance Computing School of Computer Hunan Changsha410000 China The University of Sydney Sydney AI Centre School of Computer Science Faculty of Engineering Australia

By abusing access to a well-trained classifier, model inversion (MI) attacks pose a significant threat as they can recover the original training data, leading to privacy leakage. Previous studies mitigated MI attacks by imposing regularization to reduce the dependency between input features and outputs during classifier training, a strategy known as unilateral dependency optimization. However, this strategy contradicts the objective of minimizing the supervised classification loss, which inherently seeks to maximize the dependency between input features and outputs. Consequently, there is a trade-off between improving the model's robustness against MI attacks and maintaining its classification performance. To address this issue, we propose the bilateral dependency optimization strategy (BiDO), a dual-objective approach that minimizes the dependency between input features and latent representations, while simultaneously maximizing the dependency between latent representations and labels. BiDO is remarkable for its privacy-preserving capabilities. However, models trained with BiDO exhibit diminished capabilities in out-of-distribution (OOD) detection compared to models trained with standard classification supervision. Given the open-world nature of deep learning systems, this limitation could lead to significant security risks, as encountering OOD inputs—whose label spaces do not overlap with the in-distribution (ID) data used during training-is inevitable. To address this, we leverage readily available auxiliary OOD data to enhance the OOD detection performance of models trained with BiDO. This leads to the introduction of an upgraded framework, unknown-aware BiDO (BiDO+), which mitigates both privacy and security concerns. As a highlight, with comparable model utility, BiDO-HSIC+ reduces the FPR95 by 55.02% and enhances the AUCROC by 9.52% compared to BiDO-HSIC, while also providing superior MI robustness. © 1979-2012 IEEE.

关键词： Differential privacy

来源：评论

学校读者我要写书评

暂无评论

Reinforcement Learning-Based Optimal Control for Continuous-Time Quantized Systems via Adaptive Dynamic Programming

Reinforcement Learning-Based Optimal Control for Continuous-...

引用

IEEE SSD International Multi-Conference on Systems, Signals and Devices

作者： Omar Qasem Weinan Gao Omar R. Daoud Electrical and Computer Engineering Department School of Engineering and Computing American International University Al-Jahra Kuwait State Key Laboratory of Synthetical Automation for Process Industries Northeastern University Shenyang Lianning China

ISBN: (数字)9798331542726

ISBN: (纸本)9798331542733

Ahstract- This paper presents a comprehensive investigation into the stability analysis and design conditions for quantizers to ensure the closed-loop stability of continuous-time linear quantized systems, under the conditions derived by the small-gain theorem. The study begins by deriving explicit bounds on the quantizer parameters required for maintaining system stability. Building on this foundation, an optimal controller is designed using the linear quadratic regulator (LQR) framework, providing an efficient data-driven control strategy. To further enhance the system's performance, an adaptive dynamic programming (ADP) algorithm, referred to as the hybrid iteration (HI) method, is developed. This algorithm effectively learns the optimal control policy by leveraging the trajectories of the quantized states and inputs, thereby addressing the challenges posed by quantization constraints. The proposed HI approach combines the advantages of adaptive learning and optimization, making it well-suited for continuous-time systems with limited information. The simulation results confirm that the ADP approach with the provided conditions not only stabilizes the quantized system but also achieves optimal control performance under the specified quantization conditions. This study offers valuable insights and a robust methodological framework for addressing stability and control challenges, with insights to be expanded to continuous-time nonlinear quantized systems, with potential applications in various engineering domains, such as networked systems, robotics and autonomous systems.

关键词： Quantization (signal) Adaptive systems Regulators Simulation Heuristic algorithms Optimal control Stability analysis Dynamic programming Trajectory Thermal stability

来源：评论

学校读者我要写书评

暂无评论

Dam regulation on extreme streamflow-sediment relationships along the Yangtze River Basin

引用

Science Bulletin 2025年第6期70卷 866-870页

作者： Guangyao Gao Anqi Huang Lishan Ran Dongfeng Li Hong Zhang Yadu Pokhrel Liqiang Yao Bojie Fu State Key Laboratory of Urban and Regional Ecology Research Center for Eco-Environmental SciencesChinese Academy of SciencesBeijing 100085China State Key Laboratory of Environmental Aquatic Chemistry Research Center for Eco-Environmental SciencesChinese Academy of SciencesBeijing 100085China University of Chinese Academy of Sciences Beijing 100049China Department of Geography The University of Hong KongHong Kong 999077China Key Laboratory for Water and Sediment Sciences Ministry of EducationCollege of Environmental Sciences and EngineeringPeking UniversityBeijing 100871China Department of Civil and Environmental Engineering Michigan State UniversityEast Lansing MI 48824USA Water Resources Department Yangtze River Scientific Research InstituteWuhan 430010China

Streamflow and sediment transport in river ecosystems are crit-ical processes for material movement and energy conversion,intri-cately linked to socioeconomic and riverine biophysical processes such as hydropower development,flood mitigation,reservoir silta-tion,and channel adjustment[1].Streamflow-sediment load(Q-S)relationships are fundamental to improving the assessment and prediction of these riverine processes and flux *** has substantially regulated large river systems,leading to a signif-icant decreasing trend in both streamflow and sediment load,resulting in far-reaching impacts on water quality and ecological stability[2].In the context of intensifying hydrological extremes.

关键词： prediction Yangtze River Basin extreme streamflow hydropower developmentflood mitigationreservoir sediment relationships channel adjustment streamflow sediment dam regulation streamflow sediment transport

来源：评论

学校读者我要写书评

暂无评论

Towards Event Extraction with Massive Types: LLM-based Collaborative Annotation and Partitioning Extraction

arXiv

引用

arXiv 2025年

作者： Liu, Wenxuan Li, Zixuan Bai, Long Zuo, Yuxin Xu, Daozhu Jin, Xiaolong Guo, Jiafeng Cheng, Xueqi School of Computer Science and Technology University of Chinese Academy of Sciences China Key Laboratory of Network Data Science and Technology Institute of Computing Technology Chinese Academy of Sciences China State Key Laboratory of Geo-Information Engineering Shaanxi Xi’an China

Developing a general-purpose extraction system that can extract events with massive types is a long-standing target in Event Extraction (EE). In doing so, the challenge comes from two aspects: 1) The absence of an efficient and effective annotation method. 2) The absence of a powerful extraction method can handle massive types. For the first challenge, we propose a collaborative annotation method based on Large Language Models (LLMs). Through collaboration among multiple LLMs, it first refines annotations of trigger words from distant supervision and then carries out argument annotation. Next, a voting phase consolidates the annotation preferences across different LLMs. Finally, we create the EEMT dataset, the largest EE dataset to date, featuring over 200,000 samples, 3,465 event types, and 6,297 role types. For the second challenge, we propose an LLM-based Partitioning EE method called LLM-PEE. To overcome the limited context length of LLMs, LLM-PEE first recalls candidate event types and then splits them into multiple partitions for LLMs to extract events. The results in the supervised setting show that LLM-PEE outperforms the state-of-the-art methods by 5.4% in event detection and 6.1% in argument extraction. In the zero-shot setting, LLM-PEE achieves up to 12.9% improvement compared to mainstream LLMs, demonstrating its strong generalization capabilities. Copyright © 2025, The Authors. All rights reserved.

关键词： Large datasets

来源：评论

学校读者我要写书评

暂无评论

Global method for monotone variational inequality problems on polyhedral sets

引用

OPTIMIZATION METHODS & SOFTWARE 1997年第2期7卷 111-122页

作者： Peng, JM [a] State Key Laboratory of Scientific and Engineering Computing Institute of Computational Mathematics and Scientific Computing Beijing China

In this paper, we consider the optimization method for monotone variational inequality problems on polyhedral sets. First, we consider the mixed complementarity problem based on the original problem. Then, a merit function for the mixed complementarity problem is proposed and some desirable properties of the merit function are obtained. Under certain assumptions: we show that any stationary point of the merit function is a solution of the original problem. A descent method for the optimization problem is proposed and the global convergence of the method is shown.

关键词： variational inequality problem mixed complementarity problem optimization method

来源：评论

学校读者我要写书评

暂无评论

Role of entropy in silicon carbide polytype competition

引用

Applied Physics Letters 2025年第17期126卷

作者： Jiang, Xuanyu Han, Xuefeng Pi, Xiaodong Yang, Deren Deng, Tianqi State Key Laboratory of Silicon and Advanced Semiconductor Materials School of Materials Science and Engineering Zhejiang University Hangzhou310027 China Key Laboratory of Power Semiconductor Materials and Devices of Zhejiang Province Institute of Advanced Semiconductors ZJU Hangzhou Global Scientific and Technological Innovation Center Hangzhou311215 China

Polytype control has long been a critical issue in silicon carbide single crystal growth. Empirically, the competition among different polytypes can be manipulated through temperature, carbon-to-silicon ratio, doping control, etc. However, the underlying physics remains largely elusive. Herein, we reveal that entropy difference, particularly that of the vibrational entropy and configurational entropy, is the thermodynamic origin of these polytype regulation methods. The bulk vibrational entropy difference suppresses the 3C-SiC and favors hexagonal polytypes at a higher temperature. Furthermore, the defect vibrational and configurational entropy difference arising from inequivalent carbon sites further stabilizes n-type 6H-SiC over 4H-SiC in the presence of carbon vacancies and nitrogen substitutions at elevated temperatures. These defects are reduced with high carbon chemical potentials, explaining the preference of 4H-SiC at high C/Si ratios. These entropy analyses suggest that a moderate temperature and C-rich condition are beneficial to the growth of 4H-SiC and provide physical perspectives into the thermodynamic role of phonons and defects in single crystal growth. © 2025 Author(s).

关键词： Silicon carbide

来源：评论

学校读者我要写书评

暂无评论

Movable Antenna-Assisted Integrated Sensing and Communication Systems

引用

IEEE Transactions on Wireless Communications 2025年

作者： Jiang, Chengjun Zhang, Chensi Huang, Chongwen Ge, Jianhua Niyato, Dusit Yuen, Chau Xidian University State Key Lab. of Integrated Service Networks Xi’an China Zhejiang University College of Information Science and Electronic Engineering China Xidian University State Key Laboratory of Integrated Service Networks China Zhejiang Provincial Key Laboratory of Info. Proc China Nanyang Technological University College of Computing and Data Science Singapore Nanyang Technological University School of Electrical and Electronics Engineering Singapore

Movable antennas (MAs) enhance flexibility in beamforming gain and interference suppression by adjusting position within certain areas of the transceivers. In this paper, we propose an MA-assisted integrated sensing and communication framework, wherein MAs are deployed for reconfiguring the channel array responses at both the receiver and transmitter of a base station. Then, we develop an optimization framework aimed at maximizing the sensing signal-to-interference-plus-noise-ratio (SINR) by jointly optimizing the receive beamforming vector, the transmit beamforming matrix, and the positions of MAs while meeting the minimum SINR requirement for each user. To address this nonconvex problem involving complex coupled variables, we devise an alternating optimization-based algorithm that incorporates techniques including the Charnes-Cooper transform, second-order Taylor expansion, and successive convex approximation (SCA). Specifically, the closed form of the received vector and the optimal transmit matrix can be first obtained in each iteration. Subsequently, the solutions for the positions of the transmit and receive MAs are obtained using the SCA method based on the second-order Taylor expansion. The simulation results show that the proposed scheme has significant advantages over the other baseline schemes. In particular, the proposed scheme has the ability to match the performance of the fixed position antenna scheme while utilizing fewer resources. © 2002-2012 IEEE.

关键词： Signal to noise ratio

来源：评论

学校读者我要写书评

暂无评论

IMPACT: Importance-Informed Prefetching and Caching for I/O-Bound DNN Training

引用

IEEE Transactions on Computers 2025年

作者： Chen, Weijian He, Shuibing Zhang, Ruidong Zhang, Xuechen Chen, Ping Yang, Siling Qu, Haoyang Zhan, Xuan Zhejiang University State Key Laboratory of Blockchain and Data Security China Institute of Blockchain and Data Security Hangzhou China Zhejiang Key Laboratory of Big Data Intelligent Computing China Washington State University Vancouver School of Engineering and Computer Science VancouverWA98686 United States

Fetching large amounts of DNN training data from storage systems causes high I/O latency and GPU stalls. Importance sampling can reduce data processing on GPUs while maintaining model accuracy, but current frameworks lack a prefetching and caching layer to optimize data fetches and cache management based on sample importance. This leads to unnecessary fetches, poor cache hit ratios, and random I/Os. We present ImPACT, an importance-informed prefetching and caching system, to accelerate I/O-bound DNN training. First, we propose an importance-informed prefetching technique to reduce the prefetching of unimportant data. Then, we introduce an importance-aware caching layer, partitioned into two regions: H-cache and L-cache, which store samples of high importance and low importance respectively. Rather than using recency or frequency, we manage data items in H-cache according to their corresponding sample importance. When there is a cache miss in L-cache, we use sample substitutability and dynamic packaging to improve the cache hit ratio and reduce the number of random I/Os. Our experimental results show that ImPACT has a negligible impact on training accuracy while speeding up DNN training by up to 3.5× compared to state-of-the-art prefetching and caching systems. © 1968-2012 IEEE.

关键词： Cache memory

来源：评论

学校读者我要写书评

暂无评论

A analytical kernel-based stacked pupil shift matrix decomposition method for partially coherent imaging

IEEE Transactions on Computational Imaging

引用

IEEE Transactions on Computational Imaging 2025年

作者： Zhao, Yang Xu, Yueshu Kuang, Cuifang Liu, Xu Zhejiang University State Key Laboratory of Extreme Photonics and Instrumentation College of Optical Science and Engineering Hangzhou China ZJU-Hangzhou Global Scientific and Technological Innovation Center Hangzhou China

The Sum of Coherence System (SOCS) algorithm is commonly used to calculate the aerial images of partially coherent systems in the lithography simulation. However, computing the kernels of the SOCS algorithm presents significant memory and computational challenges when dealing with the large Transmission Cross Coefficient (TCC) matrix of the Hopkins model. In this paper, we propose a method that utilizes the stacked pupil shift matrix method and analytical circle-sampling functions decomposition to avoid direct computation and storage of the TCC matrix, thereby reducing the computational consumption for obtaining the kernels. Moreover, this method demonstrates flexibility by allowing arbitrary changes to the grid size of the kernels due to its analytical nature and shows adaptability for extension to a vector model. Through numerical simulations, we demonstrate the effectiveness of our method and show that the use of analytical Fourier transforms mitigates many errors associated with discrete Fourier transform processes, further enhancing accuracy © 2015 IEEE.

关键词： Discrete Fourier transforms

来源：评论

学校读者我要写书评

暂无评论

GACL-Net:Hybrid Deep Learning Framework for Accurate Motor Imagery Classification in Stroke Rehabilitation

引用

Computers, Materials & Continua 2025年第4期83卷 517-536页

作者： Chayut Bunterngchit Laith H.Baniata Mohammad H.Baniata Ashraf ALDabbas Mohannad A.Khair Thanaphon Chearanai Sangwoo Kang State Key Laboratory of Multimodal Artificial Intelligence Systems Institute of AutomationChinese Academy of SciencesBeijing100190China School of Computing Gachon UniversitySeongnam13120Republic of Korea Computer Science Department Faculty of Information TechnologyThe World Islamic Sciences and Education UniversityAmman11947Jordan Intelligent Systems Department Faculty of Artificial IntelligenceAl-Balqa Applied UniversityAl-Salt19117Jordan IT Infrastructure Department Qatrana Cement CompanyAmman11821Jordan Division of Industrial and Logistics Engineering Technology Faculty of Engineering and TechnologyKing Mongkut’s University of Technology North BangkokRayong CampusRayong21120Thailand

Stroke is a leading cause of death and disability worldwide,significantly impairing motor and cognitive *** rehabilitation is often hindered by the heterogeneity of stroke lesions,variability in recovery patterns,and the complexity of electroencephalography(EEG)signals,which are often contaminated by *** classification of motor imagery(MI)tasks,involving the mental simulation of movements,is crucial for assessing rehabilitation strategies but is challenged by overlapping neural signatures and patient-specific *** address these challenges,this study introduces a graph-attentive convolutional long short-term memory(LSTM)network(GACL-Net),a novel hybrid deep learning model designed to improve MI classification accuracy and ***-Net incorporates multi-scale convolutional blocks for spatial feature extraction,attention fusion layers for adaptive feature prioritization,graph convolutional layers to model inter-channel dependencies,and bidi-rectional LSTM layers with attention to capture temporal *** on an open-source EEG dataset of 50 acute stroke patients performing left and right MI tasks,GACL-Net achieved 99.52%classification accuracy and 97.43%generalization accuracy under leave-one-subject-out cross-validation,outperforming existing state-of-the-art ***,its real-time processing capability,with prediction times of 33–56 ms on a T4 GPU,underscores its clinical potential for real-time neurofeedback and adaptive *** findings highlight the model’s potential for clinical applications in assessing rehabilitation effectiveness and optimizing therapy plans through precise MI classification.

关键词： Motor imagery EEG stroke rehabilitation deep learning brain-computer interface

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：