检索结果-内蒙古大学图书馆

FedCSpc: A Cross-Silo Federated Learning System with Error-Bounded Lossy Parameter Compression

IEEE Transactions on Parallel and Distributed Systems 2025年第7期36卷 1372-1386页

作者： Zhang, Zhaorui Di, Sheng Zhao, Kai Jin, Sian Tao, Dingwen Ji, Zhuoran Liu, Benben Alharthi, Khalid Ayed Cao, Jiannong Cappello, Franck The Hong Kong Polytechnic University Department of Computing Hong Kong Florida State University Department of Computer Science United States Temple University Department of Computer and Information Sciences United States Indiana University Bloomington Luddy School of Informatics Computing and Engineering United States Shandong University School of Cyber Science and Technology China The University of Hong Kong LSCM R&D Center Hong Kong University of Bisha Department of Computer Science College Of Computing And Information Technology Saudi Arabia Argonne National Laboratory United States

Cross-Silo federated learning is widely used for scaling deep neural network (DNN) training over data silos from different locations worldwide while guaranteeing data privacy. Communication has been identified as the main bottleneck when training large-scale models due to large-volume model parameters and gradient transmission across public networks with limited bandwidth. Most previous works focus on gradient compression, while limited work tries to compress parameters that can not be ignored and extremely affect communication performance during the training. To bridge this gap, we propose FedCSpc: an efficient cross-silo federated learning system with an XAI-driven adaptive parameter compression strategy for large-scale model training. Our work substantially differs from existing gradient compression techniques due to the distinct data features of gradient and parameter. The key contributions of this paper are fourfold. (1) Our designed FedCSpc proposes to compress the parameter during the training using the state-of-the-art error-bounded lossy compressor – SZ3. (2) We develop an adaptive compression error bound adjustment algorithm to guarantee the model accuracy effectively. (3) We exploit an efficient approach to utilize the idle CPU resources of clients to compress the parameters. (4) We perform a comprehensive evaluation with a wide range of models and benchmarks on a GPU cluster with 65 GPUs. Results show that FedCSpc can achieve the same model accuracy as FedAvg while reducing the data volume of parameters and gradients in communication by up to 7.39× and 288×, respectively. With 32 clients on a 4Gb size model, FedCSpc significantly outperforms FedAvg in wall-clock time in the emulated WAN environment (at the bandwidth of 1 Gbps or lower without loss of generality). © 1990-2012 IEEE.

关键词： Bandwidth compression

来源：评论

学校读者我要写书评

暂无评论

Fog Computing-based Real-Time Emotion Recognition using Physiological Signals

Fog Computing-based Real-Time Emotion Recognition using Phys...

引用

International Conference on Advanced Communication Technology (ICACT)

作者： Omur Fatmanur Erzurumluoglu Kadir Ozlem Asli Tuncay Atalay Ozgur Atalay Gökhan Ince Faculty of Computer and Informatics Engineering Istanbul Technical University Istanbul Turkey Faculty of Textile Technologies and Design Istanbul Technical University Istanbul Turkey

ISBN: (数字)9791188428137

ISBN: (纸本)9798331507602

Emotion recognition based on physiological signals has become a crucial area of research in affective computing and human-computer interaction, with applications in smart homes, workplaces, educational institutions, healthcare, and entertainment. In this study, a real-time emotion recognition system utilizing fog computing architecture was developed by considering the challenges of latency, total response time, resource usage, and security in IoT environments. The random forest machine learning model was trained with time-based statistical features by using the DREAMER dataset. Even though the model achieved an accuracy of 84.21% with 104 features, to meet real-time performance requirements, the system was optimized to calculate 24 features, maintaining a commendable accuracy of 79.70%. Extensive experiments demonstrated the superior performance of fog computing compared to edge and cloud computing in terms of latency, queuing delay, jitter, and most importantly total response time. The results highlight the proposed system’s ability to support multiple users simultaneously.

关键词： Emotion recognition Accuracy Computational modeling Smart homes Signal processing Brain modeling Real-time systems Physiology Time factors Edge computing

来源：评论

学校读者我要写书评

暂无评论

A Pipeline of Augmentation and Sequence Embedding for Classification of Imbalanced Network Traffic

arXiv

引用

arXiv 2025年

作者： Shokri, Matin Hasibi, Ramin Department of Electrical and Computer Engineering K. N. Toosi University of Technology Iran Department of Informatics University of Bergen Bergen Norway

Network Traffic Classification (NTC) is one of the most important tasks in network management. The imbalanced nature of classes on the internet presents a critical challenge in classification tasks. For example, some classes of applications are much more prevalent than others, such as HTTP. As a result, machine learning classification models do not perform well on those classes with fewer data. To address this problem, we propose a pipeline to balance the dataset and classify it using a robust and accurate embedding technique. First, we generate artificial data using Long Short-Term Memory (LSTM) networks and Kernel Density Estimation (KDE). Next, we propose replacing one-hot encoding for categorical features with a novel embedding framework based on the"Flow as a Sentence" perspective, which we name FS-Embedding. This framework treats the source and destination ports, along with the packet’s direction, as one word in a flow, then trains an embedding vector space based on these new features through the learning classification task. Finally, we compare our pipeline with the training of a Convolutional Recurrent Neural Network (CRNN) and Transformers, both with imbalanced and sampled datasets, as well as with the one-hot encoding approach. We demonstrate that the proposed augmentation pipeline, combined with FS-Embedding, increases convergence speed and leads to a significant reduction in the number of model parameters, all while maintaining the same performance in terms of accuracy. © 2025, CC BY.

关键词： Network embeddings

来源：评论

学校读者我要写书评

暂无评论

YOLOv8 to YOLO11: A Comprehensive Architecture In-depth Comparative Review

arXiv

引用

arXiv 2025年

作者： Hidayatullah, Priyanto Syakrani, Nurjannah Sholahuddin, Muhammad Rizqi Gelar, Trisna Tubagus, Refdinal Computer Engineering and Informatics Department Politeknik Negeri Bandung Kab. Bandung Barat Indonesia Stunning Vision AI Kota Cimahi Indonesia

In the field of deep learning-based computer vision, YOLO is revolutionary. With respect to deep learning models, YOLO is also the one that is evolving the most rapidly. Unfortunately, not every YOLO model possesses scholarly publications. Moreover, there exists a YOLO model that lacks a publicly accessible official architectural diagram. Naturally, this engenders challenges, such as complicating the understanding of how the model operates in practice. Furthermore, the review articles that are presently available do not delve into the specifics of each model. The objective of this study is to present a comprehensive and in-depth architecture comparison of the four most recent YOLO models, specifically YOLOv8 through YOLO11, thereby enabling readers to quickly grasp not only how each model functions, but also the distinctions between them. To analyze each YOLO version's architecture, we meticulously examined the relevant academic papers, documentation, and scrutinized the source code. The analysis reveals that while each version of YOLO has improvements in architecture and feature extraction, certain blocks remain unchanged. The lack of scholarly publications and official diagrams presents challenges for understanding the model's functionality and future enhancement. Future developers are encouraged to provide these resources. © 2025, CC BY-NC-SA.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

A Comprehensive Survey on Federated Learning Applications in Computational Mental Healthcare

引用

computer Modeling in engineering & Sciences 2025年第1期142卷 49-90页

作者： Vajratiya Vajrobol Geetika Jain Saxena Amit Pundir Sanjeev Singh Akshat Gaurav Savi Bansal Razaz Waheeb Attar Mosiur Rahman Brij B.Gupta Institute of Informatics and Communication University of DelhiDelhi110021India Maharaja Agrasen College University of DelhiDelhi110096India Computer Engineering Ronin InstituteMontclairNJ 07043USA Department of Research and Innovation Insights2TechinfoJaipur302001India University Centre for Research and Development(UCRD) Chandigarh UniversityChandigarh140413India Management Department College of Business AdministrationPrincess Nourah bint Abdulrahman UniversityRiyadh11671Saudi Arabia CCRI&Department of Computer Science and Information Engineering Asia UniversityTaichung413Taiwan Symbiosis Centre for Information Technology(SCIT) Symbiosis International UniversityPune411057India Center for Interdisciplinary Research University of Petroleum and Energy Studies(UPES)Dehradun248007India

Mental health is a significant issue worldwide,and the utilization of technology to assist mental health has seen a growing *** aims to alleviate the workload on healthcare professionals and aid *** applications have been developed to support the challenges in intelligent healthcare ***,because mental health data is sensitive,privacy concerns have *** learning has gotten some *** research reviews the studies on federated learning and mental health related to solving the issue of intelligent healthcare *** explores various dimensions of federated learning in mental health,such as datasets(their types and sources),applications categorized based on mental health symptoms,federated mental health frameworks,federated machine learning,federated deep learning,and the benefits of federated learning in mental health *** research conducts surveys to evaluate the current state of mental health applications,mainly focusing on the role of Federated Learning(FL)and related privacy and data security *** survey provides valuable insights into how these applications are emerging and evolving,specifically emphasizing FL’s impact.

关键词： Depression emotional recognition intelligent healthcare systems mental health federated learning stress detection sleep behaviour

来源：评论

学校读者我要写书评

暂无评论

IoT Integration Protocol for Enhanced Hospital Care

arXiv

引用

arXiv 2025年

作者： Zontou, Ellie Kyprioti, Antonia Dept. of Computer Engineering and Informatics University of Patras Patras26504 Greece Dept. of Medicine University of Thessaly Larisa41334 Greece

This paper introduces the "IoT Integration Protocol for Enhanced Hospital Care", a comprehensive framework designed to leverage Internet of Things (IoT) technology to enhance patient care, improve operational efficiency, and ensure data security in hospital settings. With the growing emphasis on utilizing advanced technologies in healthcare, this protocol aims to harness the potential of IoT devices to optimize patient monitoring, enable remote care, and support clinical decision-making. By integrating IoT seamlessly into nursing workflows and patient care plans, hospitals can achieve higher levels of patient-centric care and real-time data insights, leading to better treatment outcomes and resource allocation. This paper outlines the protocol’s objectives, key components, and expected benefits, while emphasizing the importance of ethical considerations and ongoing evaluation to ensure successful implementation. © 2025, CC BY-NC-ND.

关键词： Resource allocation

来源：评论

学校读者我要写书评

暂无评论

DIPLI: Deep Image Prior Lucky Imaging for Blind Astronomical Image Restoration

arXiv

引用

arXiv 2025年

作者： Singh, Suraj Batsheva, Anastasia Rogov, Oleg Y. Bouridane, Ahmed Skolkovo Institute of Science and Technology Russia College of Computing and Informatics Computer Engineering Department University of Sharjah United Arab Emirates AIRI Russia

Contemporary image restoration and super-resolution techniques effectively harness deep neural networks, markedly outperforming traditional methods. However, astrophotography presents unique challenges for deep learning due to limited training data. This work explores hybrid strategies, such as the Deep Image Prior (DIP) model, which facilitates blind training but is susceptible to overfitting, artifact generation, and instability when handling noisy images. We propose enhancements to the DIP model’s baseline performance through several advanced techniques. First, we refine the model to process multiple frames concurrently, employing the Back Projection method and the TVNet model. Next, we adopt a Markov approach incorporating Monte Carlo estimation, Langevin dynamics, and a variational input technique to achieve unbiased estimates with minimal variance and counteract overfitting effectively. Collectively, these modifications reduce the likelihood of noise learning and mitigate loss function fluctuations during training, enhancing result stability. We validated our algorithm across multiple image sets of astronomical and celestial objects, achieving performance that not only mitigates limitations of Lucky Imaging, a classical computer vision technique that remains a standard in astronomical image reconstruction but surpasses the original DIP model, state of the art transformer- and diffusion-based models, underscoring the significance of our improvements. © 2025, CC BY-NC-SA.

关键词： Markov processes

来源：评论

学校读者我要写书评

暂无评论

Design of Secure Multi-User Coded-SSK with Index Selecting Capability

引用

IEEE Transactions on Communications 2025年

作者： Hassan, Sumeyra Panayirci, Erdal Helleseth, Tor Vincent Poor, H. Kadir Has University Department of Electrical and Electronics Engineering Istanbul34083 Turkey University of Bergen Department of Informatics Bergen Norway Princeton University Department of Electrical and Computer Engineering PrincetonNJ08544 United States

We propose a new coded space shift keying (CSSK) signaling technique for multi-user (MU), multiple-input multiple-output (MIMO) communication systems incorporating physical layer security (PLS). Besides its error correction ability, the designed linear code is capable of choosing transmit antenna indices automatically and selecting the best set of antenna combinations that minimizes the bit error rate (BER). Results obtained for the single-user (SU) schemes are then extended to a general single-cell downlink MU CSSK setting. A precoder design is proposed with a maximum ratio combining (MRC) technique to eliminate the multi-user interference (MUI) entirely by taking advantage of channel state information (CSI) at the transmitter. It is shown that the same precoding provides a very effective jamming signal for the PLS against passive eavesdroppers, degrading their signal-to-interference-plus-noise ratio (SINR) severely. A closed-form expression for the achievable secrecy rates is derived and it is maximized by the proposed power allocation algorithm. Finally, it is shown analytically and by computer simulations that substantially better BER performance is achieved by each user over interference-free transmission compared to an SU transmission with a maximum likelihood (ML) detector. © 1972-2012 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

Evaluation of Multiple Apache Spark Applications using Kubernetes as a Cluster manager on Google Cloud

引用

Procedia computer Science 2025年 252卷 576-582页

作者： M. Jayanthi K. Ram Mohan Rao Vuppala Sukanya Assistant Professor Dept of Computer Science and Informatics Mahatma Gandhi University Nalgonda Professor Dept of Information Technology Vasavi College of Engineering Hyderabad Assistant Professor Dept of CSE OU Hyderabad

Big data processing frameworks demands for scalable and efficient cluster management. Apache Spark has emerged as prominent big data processing framework providing high-speed data processing and analytics capabilities for multiple applications. This paper explores the integration of Kubernetes as a cluster manager for Apache Spark applications leveraging its containerization capabilities to improve resource utilization and simplify deployment. In this paper the challenges of deploying spark applications on traditional cluster managers and showcase the advantages of adopting Kubernetes are analysed. The experimental evaluation demonstrates the benefits of Kubernetes as a cluster manager for Apache Spark framework. To execute the multiple Apache Spark applications on Kubernetes a homogenous cluster on Google Cloud is created by History bucket and service account. Finally multiple applications are executed on Google Kubernetes Engine. Output can be shown as the number of executor pods created with the performance metrics can be viewed. In conclusion, this paper compares the performance metrics such as job execution time and resource utilization with the different cluster.

关键词： Homogenous Apache Spark Kubernetes Google Kubernetes Engine Containerization

来源：评论

学校读者我要写书评

暂无评论

Designing Quantum-Safe Lattice-Based Multi-Authority CP-ABE Scheme for Blockchain-Enabled IoT-Based Consumer Healthcare Electronics

引用

IEEE Transactions on Consumer Electronics 2025年

作者： Bagchi, Prithwi Bisht, Abhishek Das, Ashok Kumar Saxena, Neetesh Hossain, M. Shamim International Institute of Information Technology Center for Security Theory and Algorithmic Research Hyderabad500 032 India Department of Computer Science and Engineering College of Informatics Korea University 145 Anam-ro Seongbuk-gu Seoul02841 Korea Republic of Cardiff University School of Computer Science and Informatics CardiffCF24 4AG United Kingdom King Saud University College of Computer and Information Sciences Department of Software Engineering Riyadh13273 Saudi Arabia

Despite edge computing reducing communication delays associated with cloud computing, privacy concerns remain a significant challenge when sharing data from edge-based consumer electronics (CE) or Internet-of-Things (IoT) devices. Ciphertext policy attribute-based encryption (CP-ABE) is a cryptographic tool that facilitates intricate and refined access control. It can deliver more flexible, secure, compact, and effective access control policies for the cloud data. In the case of a conventional CP-ABE scheme, the keys can only be issued and disseminated by a trusted central authority (CA). However, if the CA is compromised, the entire system becomes more susceptible to assaults or failures, which leads to a single-point failure. In this article, we propose a Ring-Learning with Errors (Ring-LWE)-based MA-CP-ABE scheme that is effectively resolved by the multi-authority CP-ABE (MA-CP-ABE), and it ensures the security against quantum attacks. The threshold secret sharing by Shamir and the Lagrange interpolation formula are applied in the key-generation and decryption procedures of the proposed scheme, which make it easier to segment and restore the private keys. The proposed scheme is implemented in the CE-enabled IoT-based smart healthcare applications using the blockchain technology as a secure storage. A detailed comparative study, security analysis and experimental results with the existing relevant schemes shows that the proposed scheme exhibits superior security and better efficiency as compared to other schemes, demonstrating its feasibility in practical IoT-based healthcare applications. © 2025 IEEE.

关键词： Quantum cryptography

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：