检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Hu, Zhaozhe Yin, Jia-Li Chen, Bin Lin, Luojun Chen, Bo-Hao Liu, Ximeng Fujian Province Key Laboratory of Information Security and Network Systems Fuzhou350108 China College of Computer Science and Big Data Fuzhou University Fuzhou350108 China Department of Computer Science and Engineering Yuan Ze University Taiwan

Self-ensemble adversarial training methods improve model robustness by ensembling models at different training epochs, such as model weight averaging (WA). However, previous research has shown that self-ensemble defense methods in adversarial training (AT) still suffer from robust overfitting, which severely affects the generalization performance. Empirically, in the late phases of training, the AT becomes more overfitting to the extent that the individuals for weight averaging also suffer from overfitting and produce anomalous weight values, which causes the self-ensemble model to continue to undergo robust overfitting due to the failure in removing the weight anomalies. To solve this problem, we aim to tackle the influence of outliers in the weight space in this work and propose an easy-to-operate and effective Median-Ensemble Adversarial Training (MEAT) method to solve the robust overfitting phenomenon existing in self-ensemble defense from the source by searching for the median of the historical model weights. Experimental results show that MEAT achieves the best robustness against the powerful AutoAttack and can effectively allievate the robust overfitting. We further demonstrate that most defense methods can improve robust generalization and robustness by combining with MEAT. Copyright © 2024, The Authors. All rights reserved.

关键词： Network security

来源：评论

学校读者我要写书评

暂无评论

Tackling Feature-Classifier Mismatch in Federated Learning via Prompt-Driven Feature Transformation

arXiv

引用

arXiv 2024年

作者： Wu, Xinghao Niu, Jianwei Liu, Xuefeng Shi, Mingjia Zhu, Guogang Tang, Shaojie School of Computer Science and Engineering Beihang University Beijing China Zhongguancun Laboratory Beijing China Sichuan University Sichuan China Department of Management Science and Systems University at Buffalo United States

In traditional Federated Learning approaches like FedAvg, the global model underperforms when faced with data heterogeneity. Personalized Federated Learning (PFL) enables clients to train personalized models to fit their local data distribution better. However, we surprisingly find that the feature extractor in FedAvg is superior to those in most PFL methods. More interestingly, by applying a linear transformation on local features extracted by the feature extractor to align with the classifier, FedAvg can surpass the majority of PFL methods. This suggests that the primary cause of FedAvg’s inadequate performance stems from the mismatch between the locally extracted features and the classifier. While current PFL methods mitigate this issue to some extent, their designs compromise the quality of the feature extractor, thus limiting the full potential of PFL. In this paper, we propose a new PFL framework called FedPFT to address the mismatch problem while enhancing the quality of the feature extractor. FedPFT integrates a feature transformation module, driven by personalized prompts, between the global feature extractor and classifier. In each round, clients first train prompts to transform local features to match the global classifier, followed by training model parameters. This approach can also align the training objectives of clients, reducing the impact of data heterogeneity on model collaboration. Moreover, FedPFT’s feature transformation module is highly scalable, allowing for the use of different prompts to tailor local features to various tasks. Leveraging this, we introduce a collaborative contrastive learning task to further refine feature extractor quality. Our experiments demonstrate that FedPFT outperforms state-of-the-art methods by up to 7.08%. Copyright © 2024, The Authors. All rights reserved.

关键词： Linear transformations

来源：评论

学校读者我要写书评

暂无评论

Comparative study of data preparation methods in quantum clustering algorithms 2021

Comparative study of data preparation methods in quantum clu...

引用

4th International Conference on Networking, Information systems and Security, NISS 2021

作者： Ouedrhiri, Oumayma Banouar, Oumayma Raghay, Said Elhadaj, Salah Laboratory of Computer Science Engineering and Systems FST Marrakesh Morocco Laboratory of Innovating Technologies USMBA Fes Morocco Laboratory of Applied Mathematics FST Marrakesh Morocco

ISBN: (纸本)9781450388719

More powerful and better performing, quantum algorithms offer a noticeable speedup in comparison with classical algorithms. This is due to the superposition property of quantum information. It helps to obtain algorithms that are better in terms of speed and performance. Machine learning benefits from quantum computing advantages to help time consuming algorithms run faster in the best conditions (without having to lose information). In the clustering algorithms case, the distance calculation between data points is the most resource consuming step. Thus, the use of a quantum distance is very helpful. In this paper, we test a number of quantum clustering algorithms with different data preparation methods. Experiments were done for different datasets (Iris, Wine, Breast cancer) and the comparison is based on the Clustering quality and the accuracy score. © 2021 ACM.

关键词： Quantum computing Data preparation K-means clustering Machine learning Clustering algorithms

来源：评论

学校读者我要写书评

暂无评论

A novel clinical artificial intelligence model for disease detection via retinal imaging

引用

The Innovation 2024年第2期5卷 11-12页

作者： Yidian Fu Liang Ma Sheng Wan Shengfang Ge Zhi Yang Department of Ophthalmology Ninth People’s HospitalShanghai JiaoTong University School of MedicineShanghai 200011China Shanghai Key Laboratory of Orbital Diseases and Ocular Oncology Shanghai 200011China PCA Lab Key Lab of Intelligent Perception and Systems for High-Dimensional Information of Ministry of EducationNanjing 210000China Jiangsu Key Lab of Image and Video Understanding for Social Security School of Computer Science and EngineeringNanjing University of Science and TechnologyNanjing 210000China

Artificial intelligence(AI)is an emerging field in which computerized systems are used to carry out complex tasks in place of *** AI algorithms have been developed for disease diagnosis and prediction and treatment recommendation across various clinical data types,e.g.,chest X-rays,electrocardiograms,and other radiological images.1 In ophthalmology,particularly,great progress has been made in AI systems over the past *** fundus photography(CFP)and optical coherence tomography(OCT),which are readily available in routine clinical practice,are both mainstream and useful retinal imaging modalities in *** September 2023,the 2023 Lasker-Debakey Clinical Medical Research Award was awarded to three scientists for their work on OCT for accurate retinal disease detection.

关键词： clinical routine diagnosis

来源：评论

学校读者我要写书评

暂无评论

Adaptive Constraint Partition Based Optimization Framework for Large-Scale Integer Linear Programming 37

Adaptive Constraint Partition Based Optimization Framework f...

引用

37th AAAI Conference on Artificial Intelligence, AAAI 2023

作者： Ye, Huigen Wang, Hongyan Xu, Hua Wang, Chengming Jiang, Yu State Key Laboratory of Intelligent Technology and Systems Department of Computer Science and Technology Tsinghua University Beijing100084 China School of Computer Science and Engineering Sun Yat-sen University Guangzhou China Meituan Inc. Block F&G Wangjing International R&D Park No.6 Wang Jing East Rd Chaoyang District Beijing100102 China

ISBN: (纸本)9781577358800

Integer programming problems (IPs) are challenging to be solved efficiently due to the NP-hardness, especially for large-scale IPs. To solve this type of IPs, Large neighborhood search (LNS) uses an initial feasible solution and iteratively improves it by searching a large neighborhood around the current solution. However, LNS easily steps into local optima and ignores the correlation between variables to be optimized, leading to compromised performance. This paper presents a general adaptive constraint partition-based optimization framework (ACP) for large-scale IPs that can efficiently use any existing optimization solver as a subroutine. Specifically, ACP first randomly partitions the constraints into blocks, where the number of blocks is adaptively adjusted to avoid local optima. Then, ACP uses a subroutine solver to optimize the decision variables in a randomly selected block of constraints to enhance the variable correlation. ACP is compared with LNS framework with different subroutine solvers on four IPs and a real-world IP. The experimental results demonstrate that in specified wall-clock time ACP shows better performance than SCIP and Gurobi. Copyright © 2023, Association for the Advancement of Artificial Intelligence (***). All rights reserved.

关键词： Integer programming

来源：评论

学校读者我要写书评

暂无评论

Hydrodynamic bound states of rotating microcylinders in a confining geometry

引用

Physical Review Fluids 2024年第1期9卷 014102-014102页

作者： Hanliang Guo Yi Man Hai Zhu Department of Mathematics and Computer Science Ohio Wesleyan University Delaware Ohio 43015 USA Department of Mechanics and Engineering Science at College of Engineering and State Key Laboratory for Turbulence and Complex Systems Peking University Beijing 100871 China Center for Computational Mathematics Flatiron Institute New York New York 10010 USA

Many microswimmers propel themselves by rotating microcylindrical organelles such as flagella or cilia. These cylindrical organelles almost never live in free space, yet their motions in a confining geometry can be counterintuitive. For example, one of the intriguing yet classical results in this regard is that a rotating cylinder next to a plane wall does not generate any net force in Newtonian fluids and therefore does not translate. In this work we employ analytical and numerical tools to investigate the motions of microcylinders under prescribed torques in a confining geometry. We show that a cylinder pair can form four nontrivial hydrodynamic bound states depending on the relative position within the confinement. Our analysis shows that the distinct states are the results of competing effects of the hydrodynamic interactions within the cylinder pair and between the active cylinders and the confinement.

关键词： Laminar flows Low Reynolds number swimmers

来源：评论

学校读者我要写书评

暂无评论

Coding Theorems for Repetition and Superposition Codes over Binary-Input Output-Symmetric Channels

arXiv

引用

arXiv 2024年

作者： Wang, Yixin Ma, Xiao The School of Systems Science and Engineering Sun Yat-sen University Guangzhou510006 China The Guangdong Key Laboratory of Information Security Technology School of Computer Science and Engineering Sun Yat-sen University Guangzhou510006 China

This paper is concerned with a class of low density generator matrix codes (LDGM), called repetition and superposition (RaS) codes, which have been proved to be capacity-achieving over binary-input output-symmetric (BIOS) channels in terms of bit-error rate (BER). We prove with a recently proposed framework that the RaS codes are also capacity-achieving over BIOS channels in terms of frame-error rate (FER). With this new framework, the theorem for the RaS codes can be generalized to source coding and joint source and channel coding (JSCC). In particular, we prove with this framework that the corresponding low-density parity-check (LDPC) codes, as an enlarged ensemble of quasi-cyclic LDPC (QC-LDPC) codes, can also achieve the capacity. To further improve the iterative decoding performance, we consider the convolutional RaS (Conv-RaS) code ensemble and prove it to be capacity-achieving over BIOS channels in terms of the first error event probability. The construction of Conv-RaS codes is flexible with rate (defined as the ratio of the input length to the encoding output length) ranging from less than one (typically for channel codes) to greater than one (typically for source codes), which can be implemented as a universal JSCC scheme, as confirmed by simulations. © 2024, CC BY.

关键词： Iterative decoding

来源：评论

学校读者我要写书评

暂无评论

Hierarchical Knowledge Representation of Complex Tasks Based on Dynamic Motion Primitives 7th

Hierarchical Knowledge Representation of Complex Tasks Based...

引用

7th International Conference on Cognitive systems and Information Processing, ICCSIP 2022

作者： Miao, Shengyi Zhong, Daming Miao, Runqing Sun, Fuchun Wen, Zhenkun Huang, Haiming Zhang, Xiaodong Wang, Na College of Electronics and Information Engineering Shenzhen University Shenzhen518060 China College of Modern Post Beijing University of Posts and Telecommunications Beijing100083 China Department of Computer Science and Technology Tsinghua University Beijing100083 China College of Computer Science and Software Shenzhen University Shenzhen518060 China Beijing Key Laboratory of Intelligent Space Robotic Systems Technology and Application Beijing100083 China

ISBN: (纸本)9789819906161

Current service robots without learning ability are not qualified for many complex tasks. Therefore, it is very significant to decompose the complex task into repeatable execution unit. In this paper, we propose a complex task representation method based on dynamic motion primitives, and use hierarchical knowledge graph to represent the analytic results of complex tasks. To realize the execution of complex robot manipulation tasks, we decompose the semantic tasks into the minimum motion units that can be executed by the robot and combine the multi-modal information: posture, force and robot joint parameters, obtained by the sensors. We use the knowledge graph to record the end-effector required by the robot to perform different tasks and make appropriate selection of end-effector according to different needs. Finally, Taking the long sequence complex task of service scene as an example, we use UR5 robot to verify the effectiveness and feasibility of this design. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： End effectors

来源：评论

学校读者我要写书评

暂无评论

Multiple Redundant Weighted Clustered Scheme with Dynamic Weights Adjustment for Damaged S-UAV

Multiple Redundant Weighted Clustered Scheme with Dynamic We...

引用

IEEE International Workshop on computer-Aided Modeling, Analysis, and Design of Communication Links and Networks, CAMAD

作者： Grace Khayat Constandinos X Mavromoustakis Jordi Mongay Batalla Andreas Pitsillides Department of Computer Science Mobile Systems Laboratory (MoSys Lab) University of Nicosia Nicosia Cyprus Warsaw University of Technology Warsaw Poland Department of Computer Science University of Cyprus Cyprus Department of Electrical & Electronic Engineering Science University of Johannesburg (Visiting Professor)

ISBN: (数字)9798350377644

ISBN: (纸本)9798350377651

Swarm of UAVs (S-UAVs) refers to an assembly of unmanned aerial vehicles (UAVs) working together to accomplish prearranged missions. In emergency scenarios, such as a fire, any UAV is susceptible to damage. Among the routing algorithms that S-UAVs utilize the most frequently is clustering, where UAVs are divided into clusters. Each cluster consists of a cluster head (CH) and cluster members (CM). The selection of the CH is a topic of continuous research because of its crucial significance in inter-cluster communication. We propose a clustered weighted method with dynamic weight modification and redundancy to ensure end-to-end communication despite non-functional CH. Interspace, speed, and performance indicators are combined into a weighted metric that is used to select the CH, redundant CHs, and CMs. The suggested technique optimizes UAV role selection by dynamically and autonomously adjusting the weights. Based on the results of the executed simulation, this is a promising strategy that minimizes data loss in an emergency situation and minimizes delay.

关键词： Measurement Conferences Computational modeling Redundancy Clustering algorithms Autonomous aerial vehicles Routing Routing protocols Delays Vehicle dynamics

来源：评论

学校读者我要写书评

暂无评论

Leveraging Deep Learning Models for Accurate and Reproducible Cardiac Measurements in Echocardiography

Leveraging Deep Learning Models for Accurate and Reproducibl...

引用

E-mobility, Power Control and Smart systems (ICEMPS), International Conference on

作者： Aleksandar Stojmenski Martin Mihajlov Faculty of Computer Science and Engineering Ss. Cyril and Methodius University Skopje North Macedonia Laboratory for Open Systems and Networks Jozef Stefan Institute Ljubljana Slovenia

ISBN: (数字)9798350394399

ISBN: (纸本)9798350394405

The primary metrics recorded during ultrasound sessions focused on cardiac assessment are focused on the size of the heart's features, its two chambers and pre-chambers. Presently, these measurements are predominantly conducted manually, relying on the subjective judgment of medical professionals. A sonographer takes a transthoracic echocardiogram by first applying a viscous ultrasound gel on the patient at five distinct physical locations and placing the transducer on these positions at various angles. Each angle captures slightly different perspectives of the heart called views. From each view we can make out cutouts of cardiac structures and see motion of valves and walls. This study introduces an innovative approach to cardiac assessment which delegates the evaluation task to a trained algorithm. This has the potential to expedite the measurement process and help clinicians in the interpretation of the results by proposing common cardiac issues based on pre-trained data. In the paper we provide a process for intelligent extraction of diameters along the left ventricular outflow tract (LVOT) and aortic apparatus using only the ultrasound image. We show that it is possible to successfully train a model which can extract the measurements of the diameter of the LVOT in mid systole, the diameter of the aorta annulus (AA) in mid-systole, the diameter at the aortic sinus of Valsalva (ASV) at end diastole, and the diameter of the Sino tubular junction (SJ) at end diastole. With these measures extracted, the model uses machine learning to automate the diagnostics process which can bring diagnostics closer to patients’ homes.

关键词： Decision support systems Power measurement Accuracy Ultrasonic imaging Ultrasonic variables measurement Echocardiography Volume measurement

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：