检索结果-内蒙古大学图书馆

Optimal Algorithms for Stochastic Multi-Level Compositional Optimization

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Jiang, Wei Wang, Bokun Wang, Yibo Zhang, Lijun Yang, Tianbao National Key Laboratory for Novel Software Technology Nanjing University Nanjing China Department of Computer Science The University of Iowa Iowa City United States

In this paper, we investigate the problem of stochastic multi-level compositional optimization, where the objective function is a composition of multiple smooth but possibly non-convex functions. Existing methods for solving this problem either suffer from sub-optimal sample complexities or need a huge batch size. To address these limitations, we propose a Stochastic Multi-level Variance Reduction method (SMVR), which achieves the optimal sample complexity of O(1/ϵ3 )to find an ϵ-stationary point for nonconvex objectives. Furthermore, when the objective function satisfies the convexity or Polyak-Lojasiewicz (PL) condition, we propose a stage-wise variant of SMVR and improve the sample complexity to O(1/ϵ2)for convex functions or O(1/(µϵ)) for non-convex functions satisfying the µ-PL condition. The latter result implies the same complexity for µ-strongly convex functions. To make use of adaptive learning rates, we also develop Adaptive SMVR, which achieves the same complexities but converges faster in practice. All our complexities match the lower bounds not only in terms of ϵ but also in terms of µ (for PL or strongly convex functions), without using a large batch size in each iteration. Copyright © 2022, The Authors. All rights reserved.

关键词： Stochastic systems

Low-Complexity Beamforming Design for Null Space-based Simultaneous Wireless Information and Power Transfer Systems

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Luo, Cheng Hu, Jie Xiang, Luping K., Yang Sichuan Chengdu China State Key Laboratory of Novel Software Technology Nanjing University Nanjing210008 China School of Intelligent Software and Engineering Nanjing University Suzhou Campus Suzhou China School of Computer Science and Electronic Engineering University of Essex ColchesterCO4 3SQ United Kingdom

Simultaneous wireless information and power transfer (SWIPT) is a promising technology for the upcoming sixth-generation (6G) communication networks, enabling internet of things (IoT) devices and sensors to extend their operational lifetimes. In this paper, we propose a SWIPT scheme by projecting the interference signals from both intra-wireless information transfer (WIT) and inter-wireless energy transfer (WET) into the null space, simplifying the system into a point-to-point WIT and WET problem. Upon further analysis, we confirm that dedicated energy beamforming is unnecessary. In addition, we develop a low-complexity algorithm to solve the problem efficiently, further reducing computational overhead. Numerical results validate our analysis, showing that the computational complexity is reduced by 97.5% and 99.96% for the cases of KI = KE = 2, M = 4 and KI = KE = 16, M = 64, respectively. Copyright © 2025, The Authors. All rights reserved.

关键词： Inductive power transmission

Identification of Microservices through Processed Dynamic Traces and Static Calls

学校读者我要写书评

暂无评论

Identification of Microservices through Processed Dynamic Tr...

computer science and Management technology (ICCSMT), International Conference on

作者： Sijie Wu Cheng Zhang Anhui Provincial Key Laboratory of Multimodal Cognitive Computation Anhui University School of Computer Science and Technology Anhui University State Key Lab for Novel Software Technology Nanjing University Hefei China

As the complexity of software systems continues to increase, traditional monolithic software architectures will become difficult to scale and maintain. Microservices as an emerging software architecture it advocates decentralized management, service autonomy, and distinct module boundaries. These features of microservices make software systems easy to maintain, deploy, and modify. However, microservice identification is still not well addressed due to the lack of an effective approach. The existing work on microservice identification mainly starts from the system functionality perspective and does not take into account the actual business rules, which leads to microservice candidates not conforming to the concept of microservices. To address this challenge, we propose an automated approach to microservice identification that combines dynamic and static analysis and takes into account the actual business. Our proposed approach starts with a dynamic and static analysis of the monolithic system. Once the behavioral data of the system is collected, the behavioral data will be further processed according to our proposed guidelines. Then, the proposed approach calculates the degree of the correlation between classes in a monolithic system. Finally, we propose an improved clustering algorithm based on K-means to classify the classes in a monolithic system into different microservices. The proposed approach is validated by comparison with other approaches on open source projects.

关键词：

Detecting Violations of Differential Privacy for Quantum Algorithms

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Guan, Ji Fang, Wang Huang, Mingyu Ying, Mingsheng State Key Laboratory of Computer Science Institute of Software Chinese Academy of Sciences Beijing China University of Chinese Academy of Sciences Beijing China Department of Computer Science and Technology Tsinghua University Beijing China

Quantum algorithms for solving a wide range of practical problems have been proposed in the last ten years, such as data search and analysis, product recommendation, and credit scoring. The concern about privacy and other ethical issues in quantum computing naturally rises up. In this paper, we define a formal framework for detecting violations of differential privacy for quantum algorithms. A detection algorithm is developed to verify whether a (noisy) quantum algorithm is differentially private and automatically generates bugging information when the violation of differential privacy is reported. The information consists of a pair of quantum states that violate the privacy, to illustrate the cause of the violation. Our algorithm is equipped with Tensor Networks, a highly efficient data structure, and executed both on TensorFlow Quantum and TorchQuantum which are the quantum extensions of famous machine learning platforms - TensorFlow and PyTorch, respectively. The effectiveness and efficiency of our algorithm are confirmed by the experimental results of almost all types of quantum algorithms already implemented on realistic quantum computers, including quantum supremacy algorithms (beyond the capability of classical algorithms), quantum machine learning models, quantum approximate optimization algorithms, and variational quantum eigensolvers with up to 21 quantum bits. © 2023, CC BY.

关键词： Quantum noise

Heuristic Elastic Scaling for Kubernetes Heterogeneous Microservices 21

学校读者我要写书评

暂无评论

Heuristic Elastic Scaling for Kubernetes Heterogeneous Micro...

21st International Conference on Networking, Sensing and Control, ICNSC 2024

作者： Wang, Song He, Sheng Ning, Yishuang Gao, Xiang Ding, Zhijun The Key Laboratory of Embedded System and Service Computing Ministry of Education Tongji University Shanghai201804 China The Department of Computer Science and Technology Tongji University Shanghai201804 China Kingdee Research Kingdee International Software Group Co. Ltd Shenzhen518057 China Intelligent Computing Facility Innovation Center ZhejiangLab Hangzhou30012 China

ISBN: (纸本)9798350365221

Kubernetes has become the basic platform for building cloud native applications. However, existing horizontal scaling methods based on Kubernetes have problems with resource redundancy. Furthermore, the combined horizontal and vertical scaling based on Kubernetes takes a long time. Regarding the issue above, first, we describe the execution processes of three scaling techniques and compare them in different dimensions;second, we construct a formal elastic scaling model for expansion scenarios, with the goal of minimizing the total cost;and third, we introduce a heuristic expansion algorithm after balancing resource redundancy and execution time as the heuristic information. The experiment results show that the elastic scaling algorithm proposed in this paper achieves better performance than other algorithms. © 2024 IEEE.

关键词： Cloud platforms

Enhancing Sample Utilization through Sample Adaptive Augmentation in Semi-Supervised Learning

学校读者我要写书评

暂无评论

Enhancing Sample Utilization through Sample Adaptive Augment...

International Conference on computer Vision (ICCV)

作者： Guan Gui Zhen Zhao Lei Qi Luping Zhou Lei Wang Yinghuan Shi Nanjing University National Key Laboratory for Novel Software Technology and the National Institute of Healthcare Data Science Nanjing University University of Sydney Southeast University Computer Science and Engineering Southeast University University of Wollongong

In semi-supervised learning, unlabeled samples can be utilized through augmentation and consistency regularization. However, we observed certain samples, even undergoing strong augmentation, are still correctly classified with high confidence, resulting in a loss close to zero. It indicates that these samples have been already learned well and do not provide any additional optimization benefits to the model. We refer to these samples as "naive samples". Unfortunately, existing SSL models overlook the characteristics of naive samples, and they just apply the same learning strategy to all samples. To further optimize the SSL model, we emphasize the importance of giving attention to naive samples and augmenting them in a more diverse manner. Sample adaptive augmentation (SAA) is proposed for this stated purpose and consists of two modules: 1) sample selection module; 2) sample augmentation module. Specifically, the sample selection module picks out naive samples based on historical training information at each epoch, then the naive samples will be augmented in a more diverse manner in the sample augmentation module. Thanks to the extreme ease of implementation of the above modules, SAA is advantageous for being simple and lightweight. We add SAA on top of FixMatch and FlexMatch respectively, and experiments demonstrate SAA can significantly improve the models. For example, SAA helped improve the accuracy of FixMatch from 92.50% to 94.76% and that of FlexMatch from 95.01% to 95.31% on CIFAR-10 with 40 labels. The code is available at https://***/GuanGui-nju/SAA.

关键词：

Preventing the Popular Item Embedding Based Attack in Federated Recommendations

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Zhang, Jun Li, Huan Rong, Dazhong Zhao, Yan Chen, Ke Shou, Lidan The State Key Laboratory of Blockchain and Data Security Zhejiang University Hangzhou China College of Computer Science and Technology Zhejiang University Hangzhou China Department of Computer Science Aalborg University Aalborg Denmark

Privacy concerns have led to the rise of federated recommender systems (FRS), which can create personalized models across distributed clients. However, FRS is vulnerable to poisoning attacks, where malicious users manipulate gradients to promote their target items intentionally. Existing attacks against FRS have limitations, as they depend on specific models and prior knowledge, restricting their real-world applicability. In our exploration of practical FRS vulnerabilities, we devise a model-agnostic and prior-knowledge-free attack, named PIECK (Popular Item Embedding based Attack). The core module of PIECK is popular item mining, which leverages embedding changes during FRS training to effectively identify the popular items. Built upon the core module, PIECK branches into two diverse solutions: The PIECKIPE solution employs an item popularity enhancement module, which aligns the embeddings of targeted items with the mined popular items to increase item exposure. The PIECKUEA further enhances the robustness of the attack by using a user embedding approximation module, which approximates private user embeddings using mined popular items. Upon identifying PIECK, we evaluate existing federated defense methods and find them ineffective against PIECK, as poisonous gradients inevitably overwhelm the cold target items. We then propose a novel defense method by introducing two regularization terms during user training, which constrain item popularity enhancement and user embedding approximation while preserving FRS performance. We evaluate PIECK and its defense across two base models, three real datasets, four top-tier attacks, and six general defense methods, affirming the efficacy of both PIECK and its defense. © 2025, CC BY.

关键词： Federated learning

Towards Efficient Reverse-time Migration Imaging Computation by Pipeline and Fine-grained Execution Parallelization

学校读者我要写书评

暂无评论

Towards Efficient Reverse-time Migration Imaging Computation...

IEEE International Conference on Computational science and Engineering, CSE

作者： Rong Gu Bo Li Dingjin Liu Zhaokang Wang Suhui Wangzhang Shulin Wang Haipeng Dai Yihua Huang State Key Laboratory for Novel Software Technology Nanjing University China Sinopec Geophysical Research Institute China College of Computer Science and Technology Nanjing University of Aeronautics and Astronautics China

The reverse-time migration (RTM) imaging algorithm is widely used in petroleum seismic exploration analysis. It is one of the most accurate imaging algorithms but is also computation-intensive and thus time-consuming. In this paper, we focus on improving the parallel execution performance of the reverse-time migration imaging algorithm. Firstly, we analyze the performance bottlenecks of the reverse-time migration imaging algorithm with program profiling techniques. Based on the program profiling and performance analysis, we propose three effective performance improvement strategies, including the pipeline-based iterative propagation computation, the fine-grained data compression, and the GPU memory specification-based data transmission, to eliminate the performance bottle-necks. Extensive experiments on physical clusters and real-world datasets show that the proposed pipeline-based and fine-grained parallel RTM algorithm can reduce the running time by an average of 58.42% compared with the existing solutions. In addition, the proposed algorithm has been used for over one year in the real-world production environment in Sinopec, which is one of the world's largest petroleum exploration companies.

关键词： Imaging Graphics processing units Clustering algorithms Data compression Production Companies Data communication

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

学校读者我要写书评

暂无评论

arXiv 2024年

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 Collaborative Innovation Center of Novel Software Technology and Industrialization Suzhou China School of Computer Science and Technology Soochow University Suzhou China Baidu Research Beijing China Key Lab of High Confidence Software Technologies Ministry of Education Software Institute Peking University Beijing China Department of Computer Science and Software Engineering Auburn University Auburn United States Department of Electrical and Computer Engineering North Carolina State University NC 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 paper, 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 trade-off 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, CC BY.

关键词： Federated learning