检索结果-内蒙古大学图书馆

Semantic-Aware Pseudo-labeling for Unsupervised Meta-Learning

IEEE Transactions on Pattern Analysis and Machine Intelligence 2025年第7期47卷 5475-5488页

作者： Ouyang, Tianran Dong, Xingping Ye, Mang Du, Bo Shao, Ling Shen, Jianbing Wuhan University School of Computer Science National Engineering Research Center for Multimedia Software Institute of Artificial Intelligence Hubei Key Laboratory of Multimedia and Network Communication Engineering Wuhan430072 China University of Chinese Academy of Sciences UCAS-Terminus AI Lab Beijing101408 China University of Macau State Key Laboratory of Internet of Things for Smart City Department of Computer and Information Science 314100 China

In unsupervised meta-learning, the clustering-based pseudo-labeling approach is an attractive framework, since it is model-agnostic, allowing it to synergize with supervised algorithms to learn from unlabeled data. However, the pseudo-labels suffer from clustering noise and semantic chaos problems, further impacting the effectiveness of meta-learning. In this paper, we analyze and optimize the pseudo-labeling process, including encoding and clustering, aiming to generate semantic-like pseudo-labels to narrow the gap between unsupervised and supervised meta-learning. First, during the encoding, we observe that the embedding space of existing methods lacks clustering-friendly properties, which is the primary reason for clustering noise. To address this issue, we minimize the inter-to-intra-class similarity ratio to generate clustering-friendly embedding features and validate our approach through comprehensive experiments. Then, during the clustering, we find that the semantic quality of pseudo-labels is not adequately controlled, resulting in semantic chaos of pseudo-labels. We propose a semantic-stability index to measure the semantic quality of pseudo-labels quantitatively. Based on this index, we propose the Semantic-aware Pseudo-label Reassignment mechanism to generate semantic-like pseudo-labels for all samples. Our approach is model-agnostic and can easily be integrated into existing supervised methods. To demonstrate its generalization ability, we integrate it into two representative algorithms: MAML and EP. The results on three main few-shot benchmarks clearly show that the proposed method achieves significant improvement compared to state-of-the-art models. Notably, our approach also outperforms the corresponding supervised method in three tasks. © 1979-2012 IEEE.

关键词： Latent semantic analysis

来源：评论

学校读者我要写书评

暂无评论

Efficient Federated Learning Using Dynamic Update and Adaptive Pruning with Momentum on Shared Server Data

引用

ACM Transactions on Intelligent Systems and Technology 2024年第6期15卷 1-28页

作者： Liu, Ji Jia, Juncheng Zhang, Hong Yun, Yuhui Wang, Leye Zhou, Yang Dai, Huaiyu Dou, Dejing Hithink RoyalFlush Information Network Co. Ltd. Hangzhou China School of Computer Science and Technology Soochow University Suzhou China Collaborative Innovation Center of Novel Software Technology and Industrialization Nanjing China Baidu Research Beijing China Key Lab of High Confidence Software Technologies Ministry of Education and Software Institute Peking University Beijing China Department of Computer Science and Software Engineering Auburn University AuburnAL United States Department of Electrical and Computer Engineering North Carolina State University RaleighNC United States Bedi Cloud and School of Computer Science Fudan University Beijing China

Despite achieving remarkable performance, Federated Learning (FL) encounters two important problems, i.e., low training efficiency and limited computational resources. In this article, we propose a new FL framework, i.e., FedDUMAP, with three original contributions, to leverage the shared insensitive data on the server in addition to the distributed data in edge devices so as to efficiently train a global model. First, we propose a simple dynamic server update algorithm, which takes advantage of the shared insensitive data on the server while dynamically adjusting the update steps on the server in order to speed up the convergence and improve the accuracy. Second, we propose an adaptive optimization method with the dynamic server update algorithm to exploit the global momentum on the server and each local device for superior accuracy. Third, we develop a layer-adaptive model pruning method to carry out specific pruning operations, which is adapted to the diverse features of each layer so as to attain an excellent tradeoff between effectiveness and efficiency. Our proposed FL model, FedDUMAP, combines the three original techniques and has a significantly better performance compared with baseline approaches in terms of efficiency (up to 16.9 times faster), accuracy (up to 20.4% higher), and computational cost (up to 62.6% smaller). © 2024 Copyright held by the owner/author(s). Publication rights licensed to ACM.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

Efficient and Lossless Integrity-preserving Training Scheme for High-dimensional Logistic Regression over Vertical Data

Efficient and Lossless Integrity-preserving Training Scheme ...

引用

2024 IEEE Global Communications Conference, GLOBECOM 2024

作者： Xu, Wei Zhu, Hui Liu, Ruikang Zheng, Yandong Wang, Fengwei Feng, Dengguo Xidian University School of Cyber Engineering Xi'an710126 China Institute of Software State Key Laboratory of Computer Science Beijing100190 China

ISBN: (纸本)9798350351255

Logistic regression is a widely used and efficient machine learning algorithm for data analysis. However, data is often distributed among different participants, and building high-quality models often involves data interaction, which might easily lead to sensitive information leakage. While some schemes have been proposed to address the data security issue through privacy-preserving technologies such as multi-party computation, homomorphic encryption, and trusted execution environment (TEE), they still suffer from some limitations in terms of security, efficiency, and model integrity and are not suitable for the secure outsourcing training. Therefore, in this paper, we propose an efficient and lossless integrity-preserving training for high-dimensional logistic regression over vertically partitioned data. First, we utilize a block strategy to encrypt the data in batches and outsource them to a cloud with TEE to achieve the non-interactive training. Considering TEE's limited memory, we employ random sampling technology to achieve mini-batch training within TEE, which ensures the data security. We find that the cloud might launch the active attacks to drop the model accuracy, such as lazy attack and replacement attack. To prevent the attacks, we introduce an integrity protection mechanism to verify data integrity and reduce verification frequency using a mixed-probability checking method, including fixed and random probability checking. Detailed analysis has confirmed that our scheme is secure and integrity-preserving. Extensive experiments have demonstrated that our scheme can achieve lossless, non-interactive, and efficient model training. © 2024 IEEE.

关键词： Information leakage

来源：评论

学校读者我要写书评

暂无评论

String Constraints with Regex-Counting and String-Length Solved More Efficiently 9th

String Constraints with Regex-Counting and String-Length S...

引用

9th International Symposium on Dependable software Engineering: Theories, Tools and Applications, SETTA 2023

作者： Hu, Denghang Wu, Zhilin State Key Laboratory of Computer Science Institute of Software Chinese Academy of Sciences Beijing China University of Chinese Academy of Sciences Beijing China

ISBN: (纸本)9789819986637

Regular expressions (regex for short) and string-length function are widely used in string-manipulating programs. Counting is a frequently used feature in regexes that counts the number of matchings of sub-patterns. The state-of-the-art string solvers are incapable of solving string constraints with regex-counting and string-length efficiently, especially when the counting and length bounds are large. In this work, we propose an automata-theoretic approach for solving such class of string constraints. The main idea is to symbolically model the counting operators by registers in automata instead of unfolding them explicitly, thus alleviating the state explosion problem. Moreover, the string-length function is modeled by a register as well. As a result, the satisfiability of string constraints with regex-counting and string-length is reduced to the satisfiability of linear integer arithmetic, which the off-the-shelf SMT solvers can then solve. To improve the performance further, we also propose techniques to reduce the sizes of automata. We implement the algorithms and validate our approach on 48,843 benchmark instances. The experimental results show that our approach can solve more instances than the state-of-the-art solvers, at a comparable or faster speed, especially when the counting and length bounds are large. © 2024, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Session Types with Multiple Senders Single Receiver 9th

Session Types with Multiple Senders Single Receiver

引用

9th International Symposium on Dependable software Engineering: Theories, Tools and Applications, SETTA 2023

作者： Ji, Zekun Wang, Shuling Xu, Xiong State Key Laboratory of Computer Science Institute of Software Chinese Academy of Sciences Beijing China University of Chinese Academy of Sciences Beijing China

ISBN: (纸本)9789819986637

Message passing is a fundamental element in software development, ranging from concurrent and mobile computing to distributed services, but it suffers from communication errors such as deadlocks. Session types are a typing discipline for enforcing safe structured interactions between multiple participants. However, each typed interaction is restricted to having one fixed sender and one fixed receiver. In this paper, we extend session types with existential branching types, to handle a common interaction pattern with multiple senders and a single receiver in a synchronized setting, i.e. a receiver is available to receive messages from multiple senders, and which sender actually participates in the interaction cannot be determined till execution. We build the type system with existential branching types, which retain the important properties induced by standard session types: type safety, progress (i.e. deadlock-freedom), and fidelity. We further provide a novel communication type system to guarantee progress of dynamically interleaved multiparty sessions, by abandoning the strong restrictions of existing type systems. Finally, we encode Rust multi-thread primitives in the extended session types to show its expressivity, which can be considered as an attempt to check the deadlock-freedom of Rust multi-thread programs. © 2024, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： Communications Deadlock-freedom Session types Type system

来源：评论

学校读者我要写书评

暂无评论

Three-dimensional reconstruction of industrial parts from a single image

引用

Visual Computing for Industry,Biomedicine,and Art 2024年第1期7卷 340-351页

作者： Zhenxing Xu Aizeng Wang Fei Hou Gang Zhao School of Mechanical Engineering and Automation Beihang UniversityBeijing 100191China Key Laboratory of Aeronautics Smart Manufacturing Beihang UniversityBeijing 100191China State Key Laboratory of Computer Science Institute of SoftwareChinese Academy of SciencesBeijing 100190China

This study proposes an image-based three-dimensional(3D)vector reconstruction of industrial parts that can gener-ate non-uniform rational B-splines(NURBS)surfaces with high fidelity and *** contributions of this study include three parts:first,a dataset of two-dimensional images is constructed for typical industrial parts,including hex-agonal head bolts,cylindrical gears,shoulder rings,hexagonal nuts,and cylindrical roller bearings;second,a deep learning algorithm is developed for parameter extraction of 3D industrial parts,which can determine the final 3D parameters and pose information of the reconstructed model using two new nets,CAD-ClassNet and CAD-ReconNet;and finally,a 3D vector shape reconstruction of mechanical parts is presented to generate NURBS from the obtained shape *** final reconstructed models show that the proposed approach is highly accurate,efficient,and practical.

关键词： Three-dimensional reconstruction Non-uniform rational B-splines Industrial parts Deep learning

来源：评论

学校读者我要写书评

暂无评论

A study on the impact of pre-trained model on Just-In-Time defect prediction 23

A study on the impact of pre-trained model on Just-In-Time d...

引用

23rd IEEE International Conference on software Quality, Reliability, and Security, QRS 2023

作者： Guo, Yuxiang Gao, Xiaopeng Zhang, Zhenyu Chan, W.K. Jiang, Bo Beihang University State Key Laboratory of Software Development Environment School of Computer Science and Engineering Beijing China State Key Laboratory of Computer Science Institute of Software Chinese Academy of Sciences Beijing China City University of Hong Kong Department of Computer Science Hong Kong

ISBN: (纸本)9798350319583

Previous researchers conducting Just-In-Time (JIT) defect prediction tasks have primarily focused on the performance of individual pre-trained models, without exploring the relationship between different pre-trained models as backbones. In this study, we build six models: RoBERTaJIT, CodeBERTJIT, BARTJIT, PLBARTJIT, GPT2JIT, and CodeGPTJIT, each with a distinct pre-trained model as its backbone. We systematically explore the differences and connections between these models. Specifically, we investigate the performance of the models when using Commit code and Commit message as inputs, as well as the relationship between training efficiency and model distribution among these six models. Additionally, we conduct an ablation experiment to explore the sensitivity of each model to inputs. Furthermore, we investigate how the models perform in zero-shot and few-shot scenarios. Our findings indicate that each model based on different backbones shows improvements, and when the backbone's pre-training model is similar, the training resources that need to be consumed are closer. We also observe that Commit code plays a significant role in defect detection, and different pre-trained models demonstrate better defect detection ability with a balanced dataset under few-shot scenarios. These results provide new insights for optimizing JIT defect prediction tasks using pre-trained models and highlight the factors that require more attention when constructing such models. Additionally, CodeGPTJIT and GPT2JIT achieved better performance than DeepJIT and CC2Vec on the two datasets respectively under 2000 training samples. These findings emphasize the effectiveness of transformer-based pre-trained models in JIT defect prediction tasks, especially in scenarios with limited training data. © 2023 IEEE.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

The Computational Advantage of MIP∗ Vanishes in the Presence of Noise 39

The Computational Advantage of MIP∗ Vanishes in the Presenc...

引用

39th Computational Complexity Conference, CCC 2024

作者： Dong, Yangjing Fu, Honghao Natarajan, Anand Qin, Minglong Xu, Haochen Yao, Penghui State Key Laboratory for Novel Software Technology Nanjing University China Massachusetts Institute of Technology CambridgeMA United States State Key Laboratory of Computer Science Institute of Software Chinese Academy of Sciences Beijing China Hefei National Laboratory China

ISBN: (纸本)9783959773317

The class MIP∗ of quantum multiprover interactive proof systems with entanglement is much more powerful than its classical counterpart MIP [7, 29, 28]: while MIP = NEXP, the quantum class MIP∗ is equal to RE, a class including the halting problem. This is because the provers in MIP∗ can share unbounded quantum entanglement. However, recent works [46, 47] have shown that this advantage is significantly reduced if the provers’ shared state contains noise. This paper attempts to exactly characterize the effect of noise on the computational power of quantum multiprover interactive proof systems. We investigate the quantum two-prover one-round interactive system MIP∗ [poly, O(1)], where the verifier sends polynomially many bits to the provers and the provers send back constantly many bits. We show noise completely destroys the computational advantage given by shared entanglement in this model. Specifically, we show that if the provers are allowed to share arbitrarily many EPR states, where each EPR state is affected by an arbitrarily small constant amount of noise, the resulting complexity class is equivalent to NEXP = MIP. This improves significantly on the previous best-known bound of NEEEXP (nondeterministic triply exponential time) [46]. We also show that this collapse in power is due to the noise, rather than the O(1) answer size, by showing that allowing for noiseless EPR states gives the class the full power of RE = MIP∗ [poly, poly]. Along the way, we develop two technical tools of independent interest. First, we give a new, deterministic tester for the positivity of an exponentially large matrix, provided it has a low-degree Fourier decomposition in terms of Pauli matrices. Secondly, we develop a new invariance principle for smooth matrix functions having bounded third-order Fréchet derivatives or which are Lipschitz continuous. © Yangjing Dong, Honghao Fu, Anand Natarajan, Minglong Qin, Haochen Xu, and Penghui Yao.

关键词： Fourier analysis

来源：评论

学校读者我要写书评

暂无评论

DBKEM-AACS: a distributed key escrow model in blockchain with anonymous authentication and committee selection

引用

science China(Information sciences) 2023年第3期66卷 289-290页

作者： Axin XIANG Hongfeng GAO Youliang TIAN Liang WAN College of Computer Science and Technology Guizhou University Institute of Cryptography and Data Security Guizhou University Network and Information Management Center Guizhou University State Key Laboratory of Public Big Data Guizhou University Institute of Computer Software and Theory Guizhou University

Dear editor, key security is of great practical significance and demand to guarantee the security of digital assets in the blockchain system. At present, users prefer to escrow their assets on centralized institutions, but this phenomenon has gone against the unique characteristics of decentralization and anonymity in the blockchain. Among them, the suspense incidents of assets lock or lost are enough to prove that the security of escrowed keys on the exchanges is questionable.

关键词：

来源：评论

学校读者我要写书评

暂无评论

PathActMarker:an R package for inferring pathway activity of complex diseases

引用

Frontiers of computer science 2025年第3期19卷 123-124页

作者： Xingyi LI Jun HAO Zhelin ZHAO Junming LI Xingyu LIAO Min LI Xuequn SHANG School of Computer Science Northwestern Polytechnical UniversityXi’an 710072China Research&Development Institute of Northwestern Polytechnical University in Shenzhen Shenzhen 518063China School of Software Northwestern Polytechnical UniversityXi’an 710072China Hunan Provincial Key Lab on Bioinformatics School of Computer Science and EngineeringCentral South UniversityChangsha 410083China

1 Introduction The process of complex diseases is closely linked to the disruption of key biological pathways,it is crucial to identify the dysfunctional pathways and quantify the degree of dysregulation at the indivi... 详细信息

关键词： pathway activity dysregulation complex diseases biological pathways disruption key biological pathwaysit identify dysfunctional pathways

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：