检索结果-内蒙古大学图书馆

Federated Hallucination Translation and Source-Free Regularization Adaptation in Decentralized Domain Adaptation for Foggy Scene Understanding

引用

IEEE TRANSACTIONS ON MULTIMEDIA 2025年 27卷 1601-1616页

作者： Jin, Xiating Bu, Jiajun Yu, Zhi Zhang, Hui Wang, Yaonan Zhejiang Univ Coll Software Technol Zhejiang Key Lab Accessible Percept & Intelligent Ningbo 315048 Peoples R China Zhejiang Univ Coll Comp Sci Zhejiang Key Lab Accessible Percept & Intelligent Hangzhou 310027 Peoples R China Hunan Univ Natl Engn Res Ctr Robot Visual Percept & Control T Changsha 410082 Peoples R China Hunan Univ Coll Robot Changsha 410082 Peoples R China

Semantic foggy scene understanding (SFSU) emerges a challenging task under out-of-domain distribution (OD) due to uncertain cognition caused by degraded visibility. With the strong assumption of data centralization, unsupervised domain adaptation (UDA) reduces vulnerability under OD scenario. Whereas, enlarged domain gap and growing privacy concern heavily challenge conventional UDA. Motivated by gap decomposition and data decentralization, we establish a decentralized domain adaptation (DDA) framework called Translate thEn Adapt (abbr. TEA) for privacy preservation. Our highlights lie in. (1) Regarding federated hallucination translation, a Disentanglement and Contrastive-learning based Generative Adversarial Network (abbr. DisCoGAN) is proposed to impose contrastive prior and disentangle latent space in cycle-consistent translation. To yield domain hallucination, client minimizes cross-entropy of local classifier but maximizes entropy of global model to train translator. (2) Regarding source-free regularization adaptation, a Prototypical-knowledge based Regularization Adaptation (abbr. ProRA) is presented to align joint distribution in output space. Soft adversarial learning relaxes binary label to rectify inter-domain discrepancy and inner-domain divergence. Structure clustering and entropy minimization drive intra-class features closer and inter-class features apart. Extensive experiments exhibit efficacy of our TEA which achieves 55.26% or 46.25% mIoU in adaptation from GTA5 to Foggy Cityscapes or Foggy Zurich, outperforming other DDA methods for SFSU.

关键词： Translation Adaptation models Entropy Atmospheric modeling Generative adversarial networks Bridges Semantics Robots Minimization distributed databases Decentralization federated domain adaptation source-free adaptation image translation contrastive learning prototypical knowledge foggy scene understanding

来源：评论

学校读者我要写书评

暂无评论

RAG4DS: Retrieval-Augmented Generation for Data Spaces-A Unified Lifecycle, Challenges, and Opportunities

引用

IEEE ACCESS 2025年 13卷 39510-39522页

作者： Al-Qatf, Majjed Haque, Rafiqul Alsamhi, Saeed Hamood Buosi, Samuele Razzaq, Muhammad Asif Timilsina, Mohan Hawbani, Ammar Curry, Edward Univ Galway Insight Ctr Data Analyt Galway H91TK33 Ireland Korea Univ Coll Informat Dept Comp Sci & Engn Seoul 02841 South Korea IBB Univ Fac Engn Elect Engn Dept Ibb Yemen Shenyang Aerosp Univ Sch Comp Sci Shenyang 110100 Peoples R China

Retrieval-Augmented Generation (RAG) has gained significant attention from many researchers as an effective solution to address the hallucination issue of Foundational Models (FMs), particularly Large Language Models (LLMs). Although the RAG framework is considered a successful approach for enhancing LLMs by providing a suitable retrieval mechanism to obtain appropriate external knowledge, it still has limitations in acquiring high-quality knowledge from diverse data sources. The complementary integration of RAG and data spaces is proposed to exploit RAG's capabilities within data spaces. Data spaces provide RAG with the ability to obtain diverse and high quality data sources from several data providers under secure data-sharing mechanisms and direct data exchange negotiations. At the same time, RAG enhances the support services of data spaces. In this paper, we present a high-level architecture for RAG data space models (RAG-DSMs) with a unified lifecycle for RAG and data spaces, highlight the possible challenges of the proposed integration while presenting potential opportunities. Moreover, we present two use cases for leveraging RAG-DSMs in the mobility and health domains.

关键词： Frequency modulation Data models Soft sensors distributed databases Retrieval augmented generation Foundation models Data privacy Technological innovation Sustainable development Real-time systems Artificial intelligence foundational models retrieval-augmented generation data spaces lifecycle

来源：评论

学校读者我要写书评

暂无评论

Order Optimal Bounds for One-Shot Federated Learning Over Non-Convex Loss Functions

引用

IEEE TRANSACTIONS ON INFORMATION THEORY 2024年第4期70卷 2807-2830页

作者： Sharifnassab, Arsalan Salehkaleybar, Saber Golestani, S. Jamaloddin Univ Alberta Comp Sci Dept Reinforcement Learning & Artificial Intelligence R Edmonton AB T6G 2E8 Canada Leiden Univ Leiden Inst Adv Comp Sci LIACS NL-2311 EZ Leiden Netherlands Sharif Univ Technol Dept Elect Engn Tehran 113659466 Iran

We consider the problem of federated learning in a one-shot setting in which there are m machines, each observing n sample functions from an unknown distribution on non-convex loss functions. Let F:[-1, 1](d )-> R be the expected loss function with respect to this unknown distribution. The goal is to find an estimate of the minimizer of F . Based on its observations, each machine generates a signal of bounded length B and sends it to a server. The server collects signals of all machines and outputs an estimate of the minimizer of F . We show that the expected loss of any algorithm is lower bounded by max(1 root n-(mB)(1/d)), 1 root mn) , up to a logarithmic factor. We then prove that this lower bound is order optimal in m and n by presenting a distributed learning algorithm, called Multi-Resolution Estimator for Non-Convex loss function (MRE-NC), whose expected loss matches the lower bound for large mn up to polylogarithmic factors.

关键词： Servers Federated learning Manganese Computational modeling distributed databases Machinery Training distributed learning communication efficiency non-convex optimization

来源：评论

学校读者我要写书评

暂无评论

Compression Analysis for BlueField-2/-3 Data Processing Units: Lossy and Lossless Perspectives

引用

IEEE MICRO 2024年第2期44卷 8-19页

作者： Li, Yuke Kashyap, Arjun Guo, Yanfei Lu, Xiaoyi Univ Calif Merced Dept Comp Sci & Engn Parallel & Distributed Syst Lab Merced CA 95343 USA Argonne Natl Lab Lemont IL 60439 USA Univ Calif Merced Dept Comp Sci & Engn Merced CA 95343 USA

A data processing unit (DPU) with programmable smart network interface card containing system-on-chip (SoC) cores is now a valuable addition to the host CPU, finding use in high-performance computing (HPC) and data center clusters for its advanced features, notably, a hardware-based data compression engine (C-engine). With the convergence of big data, HPC, and machine learning, data volumes burden communication and storage, making efficient compression vital. This positions DPUs as tools to accelerate compression workloads and enhance data-intensive applications. This article characterizes lossy (e.g., SZ3) and lossless (e.g., DEFLATE, lz4, and zlib) compression algorithms using seven real-world datasets on Nvidia BlueField-2/-3 DPUs. We explore the potential opportunities for offloading these compression workloads from the host. Our findings demonstrate that the C-engine within the DPU can achieve up to 26.8x speedup compared to its SoC core. We also provide insights on harnessing BlueField for compression, presenting seven crucial takeaways to steer future compression research with DPUs.

关键词： Engines Task analysis Data compression Throughput Memory management Hardware acceleration distributed databases

来源：评论

学校读者我要写书评

暂无评论

ISFL: Federated Learning for Non-i.i.d. Data With Local Importance Sampling

引用

IEEE INTERNET OF THINGS JOURNAL 2024年第16期11卷 27448-27462页

作者： Zhu, Zheqi Shi, Yuchen Fan, Pingyi Peng, Chenghui Letaief, Khaled B. Tsinghua Univ Dept Elect Engn Beijing 10084 Peoples R China Huawei Technol Wireless Technol Lab Shenzhen 518129 Peoples R China Hong Kong Univ Sci & Technol Dept Elect & Comp Engn Hong Kong Peoples R China

As a promising learning paradigm integrating computation and communication, federated learning (FL) proceeds the local training and the periodic sharing from distributed clients. Due to the non-i.i.d. data distribution on clients, FL model suffers from the gradient diversity, poor performance, bad convergence, etc. In this work, we aim to tackle this key issue by adopting importance sampling (IS) for local training. We propose importance sampling FL(ISFL), an explicit framework with theoretical guarantees. First, we derive the convergence theorem of ISFL to involve the effects of local IS. Then, we formulate the problem of selecting optimal IS weights and obtain the theoretical solutions. We also employ a water-filling method to calculate the IS weights and develop the ISFL algorithms. The experimental results on CIFAR-10 fit the proposed theorems well and verify that ISFL reaps better performance, convergence, sampling efficiency, as well as explainability on the non-i.i.d. data. To the best of our knowledge, ISFL is the first non-i.i.d. FL solution from the local sampling aspect which exhibits theoretical compatibility with neural network models. Furthermore, as a local sampling approach, ISFL can be easily migrated into the other emerging FL frameworks.

关键词： Training Monte Carlo methods Data models Convergence Federated learning distributed databases Deep learning Federated learning (FL) importance sampling (IS) non i.i.d data water-filling optimization

来源：评论

学校读者我要写书评

暂无评论

An Adaptive Federated Fuzzy C-Means Clustering With Nonindependently and Identically distributed Data

引用

IEEE TRANSACTIONS ON SYSTEMS MAN CYBERNETICS-SYSTEMS 2025年第6期55卷 4015-4028页

作者： Li, Longmei Lu, Wei Pedrycz, Witold Dalian Univ Technol Sch Control Sci & Engn Dalian 116024 Liaoning Peoples R China Univ Alberta Dept Elect & Comp Engn Edmonton AB T6G 2R3 Canada Macau Univ Sci & Technol Inst Syst Engn Taipa Peoples R China Istinye Univ Res Ctr Performance & Prod Anal TR-34396 Istanbul Turkiye

Federated Fuzzy C-Means (FCM) has received considerable attention due to the increasing need for privacy-conscious data analysis across diverse domains and sources in many real-world applications. Recent developments in federated FCM, however, are still in their infancy and largely unexplored. These methods struggle to handle nonindependent and identically distributed (non-iid) data. Moreover, critical hyperparameters, such as the number of iterations for local updates, are typically set manually, which can significantly affect the performance of federated clustering. To address these challenges, we introduce an Adaptive Federated FCM with an auxiliary model, named AF-FCM. In this approach, prior information from the auxiliary model, along with a proximal term in the local objective, mitigates the effects of the non-iid environment, enhancing both model robustness and effectiveness. Critical hyperparameters are adaptively adjusted using a proposed adaptive particle swarm optimization (APSO) algorithm, guided by a carefully designed fitness function. Within APSO, a nonlinear regression function adjusts the inertia weight, reducing the risk of convergence to local optima. In AF-FCM, global prototypes are refined using momentum gradient descent (MGD). Numerical experiments highlight the effectiveness and generalization performance of AF-FCM across various conditions, including heterogeneity variations, the number of clients, and the number of clusters. Comparative analysis against state-of-the-art federated clustering baseline methods validates the competitive performance of AF-FCM.

关键词： Prototypes Servers Adaptation models distributed databases Data privacy Training Linear programming Fuzzy logic Computational modeling Protocols distributed learning federated learning (FL) fuzzy clustering nonindependent and identically distributed (non-iid)

来源：评论

学校读者我要写书评

暂无评论

An Efficient Sequential Decentralized Federated Progressive Channel Pruning Strategy for Smart Grid Electricity Theft Detection

引用

IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS 2025年第3期21卷 2393-2402页

作者： Guo, Fanghong Li, Shengwei Yang, Hao Dong, Chen Chen, Yifang Li, Guoqi Zhejiang Univ Technol Coll Informat Engn Hangzhou 310023 Peoples R China Hangzhou Power Supply Co State Grid Zhejiang Elect Power Co Ltd Hangzhou 310016 Peoples R China Chinese Acad Sci Inst Automat Beijing 100190 Peoples R China

This article aims to develop a lightweight, decentralized federated learning (FL)-based strategy for electricity theft detection (ETD). Different from most of the existing ETD solutions, which typically deploy centralized deep learning models, our proposed method utilizes well-pruned lightweight networks and operates in a completely decentralized manner while maintaining the performance of the ETD model. Specifically, to protect data privacy, a novel sequential decentralized FL (SDFL) framework was designed, eliminating the centralized parameter aggregation node in traditional FL. Each client communicates model parameters only with its neighbors and trains its model locally. In addition, to facilitate deployment on edge devices, model pruning techniques are integrated with the sequential transmission characteristics of the SDFL framework. A progressive channel pruning technique is proposed, gradually reducing the number of model channels during training to promote model compression and simplify field deployment. Experiments demonstrate that our strategy compressed the model floating point operations from 18.32 to 3.60M and reduced the number of parameters from 8.61 to 3.47M, while protecting user privacy, and maintaining good performance. Deployment results on the edge devices, i.e., Raspberry Pi, indicate that our proposed strategy reduces the model inference time from 329.35 to 141.50 s, enhancing the detection efficiency by 57.04%.

关键词： Data models Computational modeling Training Servers Biological system modeling Data privacy Electricity Adaptation models Costs distributed databases Decentralized federated learning (FL) electricity theft detection (ETD) model pruning privacy protection smart grid

来源：评论

学校读者我要写书评

暂无评论

Task-Aware Service Placement for distributed Learning in Wireless Edge Networks

引用

IEEE TRANSACTIONS ON PARALLEL AND distributed SYSTEMS 2025年第4期36卷 731-744页

作者： Cong, Rong Zhao, Zhiwei Wang, Mengfan Min, Geyong Liu, Jiangshu Mo, Jiwei Univ Elect Sci & Technol China UESTC Chengdu 610051 Sichuan Peoples R China Univ Elect Sci & Technol China Coll Comp Sci & Engn Chengdu 610051 Sichuan Peoples R China Univ Exeter Coll Engn Math & Phys Sci Exeter EX4 4PY England

Machine learning has been a driving force in the evolution of tremendous computing services and applications in the past decade. Traditional learning systems rely on centralized training and inference, which poses serious privacy and security concerns. To solve this problem, distributed learning over wireless edge networks (DLWENs) emerges as a trending solution and has attracted increasing research interests. In DLWENs, corresponding services need to be placed onto the edge servers to process the distributed tasks. Apparently, different placement of training services can significantly affect the performance of all distributed learning tasks. In this article, we propose TASP, a task-aware service placement scheme for distributed learning in wireless edge networks. By carefully considering the structures (directed acyclic graphs) of the distributed learning tasks, the fine-grained task requests and inter-task dependencies are incorporated into the placement strategies to realize the parallel computation of learning services. We also exploit queuing theory to characterize the dynamics caused by task uncertainties. Extensive experiments based on the Alibaba ML dataset show that, compared to the state-of-the-art schemes, the proposed work reduces the overall delay of distributed learning tasks by 38.6% on average.

关键词： Servers Distance learning Computer aided instruction Wireless communication Training Edge computing distributed databases Delays Data models Computational modeling distributed learning edge computing service placement task-aware

来源：评论

学校读者我要写书评

暂无评论

Identification and Alleviation of False Data Injection Within the Cyber Layer of an Enhanced distributed Secondary Control in DC Islanded Microgrids

引用

IEEE ACCESS 2025年 13卷 48605-48624页

作者： Mosaad, Nada Abdel-Rahim, Omar Rohouma, Wesam Abdelkader, Sobhy Mohamed Egypt Japan Univ Sci & Technol E JUST Elect Power Engn New Borg El Arab City 21934 Alexandria Egypt Aswan Univ Fac Engn Elect Engn Dept Aswan 81542 Egypt Univ Doha Sci & Technol Elect Engn Dept Doha Qatar Mansoura Univ Fac Engn Elect Engn Dept Mansoura 35516 Egypt

As a result of the integration of communication links within secondary control to circumvent the drawbacks associated with centralized control, microgrids have evolved into cyber-physical systems which are susceptible to cybersecurity risks. This research addresses the repercussions of injecting fake data into the cyber layer within the framework of distributed secondary control in a DC-islanded microgrid. Kullback-Leibler divergence, also referred to as relative entropy, serves as a statistical metric between two probability distributions employed to compute a specific parameter tasked with mitigating the impact on the control system. This parameter comprises the multiplication of two components. The first component is an internal component derived from neighboring entities and provides insight into their security status, indicating whether they have been subjected to attacks. The second component is an external component computed locally and verifies the accuracy of the status received from each neighbor, ensuring its authenticity. Two medium voltage DC microgrids were examined through MATLAB Simulink simulations to corroborate the efficiency of the suggested approach in identifying and mitigating the consequences of false data injection. The results demonstrated that the cybersecurity threat's influence could be proficiently detected and alleviated in both microgrid configurations, while maintaining the effectiveness of secondary control functions, which are, preserving voltage regulation and precise current distribution. Finally, one of the Microgrids was implemented using OPAL-RT, allowing for the observation of output voltages and currents on a physical hardware oscilloscope.

关键词： Microgrids distributed databases distributed power generation Voltage control Decentralized control Mathematical models MATLAB Stability analysis Renewable energy sources Power system stability distributed secondary control false data injection Kullback-Leibler divergence MATLAB Simulink OPAL-RT

来源：评论

学校读者我要写书评

暂无评论

A Dual Inexact Nonsmooth Newton Method for distributed Optimization

引用

IEEE TRANSACTIONS ON SIGNAL PROCESSING 2025年 73卷 188-203页

作者： Niu, Dunbiao Hong, Yiguang Song, Enbin Tongji Univ Coll Elect & Informat Engn Dept Control Sci & Engn Natl Key Lab Autonomous Intelligent Unmanned Syst Shanghai 201804 Peoples R China Sichuan Univ Coll Math Chengdu 610065 Sichuan Peoples R China

In this paper, we propose a novel dual inexact nonsmooth Newton (DINN) method for solving a distributed optimization problem, which aims to minimize a sum of cost functions located among agents by communicating only with their neighboring agents over a network. Our method is based on the Lagrange dual of an appropriately formulated primal problem created by introducing local variables for each agent and enforcing a consensus constraint among these variables. Due to the decomposed structure of the dual problem, the DINN method guarantees a superlinear (or even quadratic) convergence rate for both the primal and dual iteration sequences, achieving the same convergence rate as its centralized counterpart. Furthermore, by exploiting the special structure of the dual generalized Hessian, we design a distributed iterative method based on Nesterov's acceleration technique to approximate the dual Newton direction with suitable precision. Moreover, in contrast to existing second-order methods, the DINN method relaxes the requirement for the objective function to be twice continuously differentiable by using the linear Newton approximation of its gradient. This expands the potential applications of distributed Newton methods. Numerical experiments demonstrate that the DINN method outperforms the current state-of-the-art distributed optimization methods.

关键词： Convergence Jacobian matrices Vectors Newton method Linear programming Costs Signal processing algorithms Approximation algorithms distributed databases Cost function Inexact nonsmooth Newton method distributed optimization superlinear convergence rate communication cost

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：