检索结果-内蒙古大学图书馆

Fed RDP: A Robust Federated Learning Framework for Multi Party Decision-Making

IEEE Transactions on Network and Service Management 2025年

作者： An, Jian Tang, Siyu Ping, Ruyuan Li, Ran He, Xin Jin, Xiaolong Xi’an Jiaotong University School of Computer Science and Technology China Shaanxi Province Key Laboratory of Computer Network Xi’an Jiaotong University Xi’an710049 China Xi’an Jiaotong University School of Computer Science and Technology Xi’an710049 China Henan University School of Software Kaifeng475001 China Rocket Force University of Engineering School of Combat Supporting Xi’an710000 China

In the field of intelligent manufacturing, the vast and heterogeneous nature of data across different departments poses significant challenges for collaborative decision-making. Direct data sharing often leads to severe privacy and security concerns. Although federated learning offers a promising solution, issues such as participant selection and privacy protection remain inadequately addressed. During model training, it is crucial to minimize the influence of low-quality participants and prevent the inference of sensitive information through gradient analysis, which could threaten model performance or data privacy. To address these challenges, this paper proposes a robust federated learning model, Fed-RDP, based on participant contribution evaluation and personalized differential privacy. The model leverages blockchain technology for secure data flow and storage, implemented through smart contracts. Historical parameters are submitted to evaluate contributions based on the concept of the least core, with real-time updates to reputation scores. When participants upload their local models, the scores are used as weights to aggregate and update the global model. Additionally, personalized differential noise is added to the uploaded gradients based on participant scores, preserving privacy while maximizing the utility of the data. Experimental results demonstrate that this approach effectively identifies low-quality participants, optimizes evaluation time, and protects privacy through personalized differential noise. © 2004-2012 IEEE.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

Intelligent photonic crystal-based optical sensor for accurate glycerol-water mixture measurement using artificial neural networks

引用

Optical Materials Express 2025年第7期15卷 1710-1729页

作者： Saeed Roshani Salah I Yahya Pouya Karami Muhammad Akmal Chaudhary Maher Assaad Fariborz Parin Fawwaz Hazzazi Fawnizu Azmadi Hussin Sobhan Roshani Department of Electrical Engineering Ker.C. Islamic Azad University Kermanshah Iran Department of Software Engineering Faculty of Engineering Koya University Koya KOY45 Iraq Department of Computer Technology Engineering College of Technical Engineering Al-Hadba University Mosul Iraq Department of Electrical and Computer Engineering College of Engineering and Information Technology Ajman University Ajman 346 United Arab Emirates Department of Electrical Engineering College of Engineering Prince Sattam bin Abdulaziz University Al-Kharj 11492 Saudi Arabia Department of Electrical and Electronic Engineering Universiti Teknologi PETRONAS Seri Iskandar 32610 Malaysia

This paper proposes an intelligent photonic crystal-based optical sensor designed for the first time to accurately measure glycerol-water concentration and temperature. The proposed sensor features a novel two-dimensional (2D) photonic crystal structure with an optimized waveguide configuration to enhance refractive index sensitivity. The sensor structure does not include defect rods, which simplifies fabrication and enhances stability. By using the unique optical properties of photonic crystals and the artificial neural network (ANN), the proposed design ensures high precision and stability in detecting changes in the glycerol concentration. The performance of the sensor was evaluated based on sensitivity, detection limit (DL), figure of merit (FOM), and quality factor (Q-F) across different temperatures and glycerol concentrations. The optical response of the sensor was numerically analyzed and simulated using the finite-difference time-domain (FDTD) method. Then, a feedforward ANN model was developed and trained to predict glycerol concentration and temperature from the output spectral data, enabling intelligent and real-time analysis. The results demonstrate that the proposed sensor achieves high sensitivity (up to 89.9 nm/RIU), a low detection limit (0.0003–0.0010 RIU −1 ), and an excellent Q-factor (5233), making it a highly effective solution for refractive index sensing. Overall, the findings confirm that the proposed photonic crystal sensor, enhanced with ANN-based intelligent analysis, offers high accuracy, stability, and fast response, making it suitable for biomedical, pharmaceutical, and industrial applications where precise glycerol concentration measurements are required.

关键词： Effective refractive index Nematic liquid crystals Neural networks Photonic crystal sensors Photonic crystals Sensor performance

来源：评论

学校读者我要写书评

暂无评论

A Taxonomy of AI-Based Assessment Educational Technologies

A Taxonomy of AI-Based Assessment Educational Technologies

引用

International Conference on Information and Communication Systems (ICICS)

作者： Mahmoud M. Hammad Mohammed Al-Refai Wafaa Musallam Sajida Musleh Esra’a Faouri Software Engineering Computer and Information Technology Center for E-Learning and Open Educational Resources Jordan University of Science and Technology Irbid Jordan Computer Science Computer and Information Technology Center for E-Learning and Open Educational Resources Jordan University of Science and Technology Irbid Jordan Center for E-Learning and Open Educational Resources Jordan University of Science and Technology Irbid Jordan

ISBN: (数字)9798331540760

ISBN: (纸本)9798331540777

Artificial Intelligence (AI) is increasingly applied across various domains, including education, where it enhances numerous aspects of the learning process, from course design to assessment. Despite its benefits in efficiency, scalability, and consistency, AI in education is applied in different learning and educational stages. This paper focuses on the use of AI in the assessment stage. To that end, this paper proposes a taxonomy of AI-based learner assessment educational technologies (EduTech) from both research and industrial perspectives. The taxonomy provides a comprehensive understanding and identifies gaps in the field. Using the PRISMA framework, we systematically review related research papers and tools.

关键词： Reviews Communication systems Scalability Taxonomy Educational technology Learning (artificial intelligence)

来源：评论

学校读者我要写书评

暂无评论

Revolutionizing Image Recognition: Next-Generation CNN Architectures for Handwritten Digits and Objects

Revolutionizing Image Recognition: Next-Generation CNN Archi...

引用

IEEE Symposium on Wireless technology and Applications (ISWTA)

作者： Md Nurul Absur Kazi Fahim Ahmad Nasif Sourya Saha Sifat Nawrin Nova Department of Computer Science City University of New York New York USA College of Computing and Software Engineering Kennesaw State University Georgia USA Department of Computer Science Chalmers University of Technology Gothenburg Sweden

ISBN: (数字)9798350351354

ISBN: (纸本)9798350351361

This study addresses the pressing need for computer systems to interpret digital media images with a level of sophistication comparable to human visual perception. By leveraging Convolutional Neural Networks (CNNs), we introduce two innovative architectures tailored to distinct datasets: the MNIST handwritten digit dataset and the Fashion MNIST dataset. Unlike traditional machine learning methods such as Support Vector Machines (SVM) and Random Forests, our customized CNN models remarkably enhance image attribute comprehension and recognition accuracy. Specifically, the model developed for the MNIST dataset achieved an unprecedented accuracy of 98.71% without any bias, while the Fashion MNIST model reached 91.39%, marking significant advancements over conventional algorithms without any bias. This research showcases the superior efficiency of CNNs in processing and understanding digital images. It underscores the potential of deep learning technologies in bridging the gap between computational systems and human-like visual recognition. Through meticulous experimentation and analysis, we illustrate how deep CNNs require less preparatory work than other image-processing algorithms, setting a new benchmark in computer vision.

关键词： Support vector machines Handwriting recognition Visualization Accuracy Image recognition Computational modeling computer architecture

来源：评论

学校读者我要写书评

暂无评论

Research on collaborative distance English teaching system based on mobile cloud computing

引用

International Journal of Industrial and Systems engineering 2021年第3期39卷 313-329页

作者： Shi, Xiaozhen Hou, Lei Wang, Dunn School of Foreign Languages Xuzhou University of Technology Xuzhou221000 China School of Information and Electronic Engineering Xuzhou University of Technology Xuzhou221000 China School of Computer Science and Software Engineering University of Southampton SouthamptonSO17-1BJ United Kingdom

Aiming at the problems of low success rate and poor robustness of traditional distance English teaching system, this paper proposes a new collaborative distance English teaching system based on mobile cloud computing. The hardware of the system adopts USB-to-JTAG device to download files and adopts reverse principle to upload English teaching resources. The four-wire signal is purified and denoised by phase-locked loop. When the system fails, reset circuit is used to deal with it in time. The system software adopts the structure of the Yangtze River basin, establishes the virtual teaching place and video voice teaching method, and realises the remote synchronous teaching. The experimental results show that the success rate and robustness of distance cooperative English teaching system based on mobile cloud computing are 98.2% and 94.5%, which are significantly higher than traditional methods. This method has good reliability and is more suitable for English teaching system. Copyright © 2021 Inderscience Enterprises Ltd.

关键词： Teaching

来源：评论

学校读者我要写书评

暂无评论

Articulated Object Manipulation using Online Axis Estimation with SAM2-Based Tracking

arXiv

引用

arXiv 2024年

作者： Wang, Xi Chen, Tianxing Yu, Qiaojun Xu, Tianling Chen, Zanxin Fu, Yiting He, Ziqi Lu, Cewu Mu, Yao Luo, Ping Department of Computer Science The University of Hong Kong Hong Kong College of Computer Science and Software Engineering Shenzhen University Shenzhen China Shanghai Jiaotong University Shanghai China Southern University of Science and Technology China

Articulated object manipulation requires precise modeling, where understanding the 3D motion constraints of individual articulated components is crucial. Prior research has leveraged interactive perception to facilitate manipulation;however, open-loop approaches often fail to account for the interaction dynamics, limiting their effectiveness. To overcome this limitation, we propose a closed-loop pipeline that integrates interactive perception with online axis estimation and optimization. Our method processes sequentially segmented 3D point clouds of manipulated object parts, continuously refining the motion model throughout the manipulation process. Specifically, our approach builds upon any interactive perception technique, inducing slight object movements to capture point cloud frames of the evolving dynamic scene. These point clouds are segmented using the Segment Anything Model 2 (SAM2) [Ravi et al.], where the moving object part is accurately masked for online axis estimation. The estimated axis then guides subsequent robotic motion planning, ensuring precise and adaptive manipulation. Experimental results in simulated environments demonstrate that our method outperforms baseline approaches, particularly in tasks requiring precise axis-based control. By integrating real-time perception with online optimization, our framework significantly improves both the accuracy and efficiency of articulated object manipulation. Copyright © 2024, The Authors. All rights reserved.

关键词： Robot programming

来源：评论

学校读者我要写书评

暂无评论

Machine Learning-based Multi-path Reliable and Energy-efficient Routing Protocol for Underwater Wireless Sensor Networks

Machine Learning-based Multi-path Reliable and Energy-effici...

引用

Frontiers of Information technology (FIT)

作者： Zahid Ullah Khan Muhammad Aman Wazir Ur Rahman Faran Khan Tooba Jamil Rifhat Hashim College of Underwater Acoustic Engineering Harbin Engineering University Harbin China Department of Software Engineering University of Science and Technology Bannu Pakistan Department of Computer Science Barani Institute of Information and Technology Rawal Pindi Pakistan

The research on applications for Underwater Wireless Sensor Networks (UWSNs) including the investigation of submerged resources, gathering oceanographic data, conducting operational surveillance, and prevention from natural disasters is rapidly increasing. Traditional wireless sensor networks (TWSNs) are distinct from UWSNs. Routing protocols of TWSNs are also distinct from UWSNs routing protocols in terms of energy efficiency. Most of routing protocols have been developed to extract data from the ocean floor to the surface of waters. There are many challenges for indigenous designing of an efficient routing protocol for UWSNs, such as, design of reliable path for forwarding communicating data packets, managing the movement of nodes, configuration of sensor nodes, removing void communicating nodes and increasing the power efficiency of the system. This research focuses on designing a novel Machine learning-based multi-path reliable and energy-efficient routing protocol (M2RE2RP) for UWSNs. It will select an efficient path among the existing communication paths used by sensor nodes, which will ultimately increase the lifespan of the network. The simulation experiment results dictate that the proposed routing model M2RE2RP achieves better performance in terms of network throughput, end-to-end latency, total energy consumption and network lifetime.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Traffic Prediction Using Lightweight Large Model in UAV-Assisted Mobile Computing for Time-Critical Consumer Electronics

引用

IEEE Transactions on Consumer Electronics 2025年

作者： Jiang, Yunzhe Liu, Yinqiu Cui, Huajun Liu, Jinghuan Mao, Yingchi He, Xiaoming University of Electronic Science and Technology of China College of Communication and Information Engineering Chengdu China Nanyang Technological University School of Computer Science and Engineering Singapore PowerChina Beijing Engineering Corporation Limited Digital Intelligence Research Institute Beijing China Hohai University College of Computer Science and Software Engineering Nanjing China Nanjing University of Posts and Telecommunications College of Internet of Things Nanjing China

Unmanned Aerial Vehicles (UAVs) have emerged as an important category of consumer electronics, offering versatile applications and enhanced operational efficiency. In the context of Mobile Edge Computing (MEC), UAVs can improve task execution and service quality by providing dynamic allocation of computational and data processing resources at the network edge. However, one critical issue in UAV-assisted MEC is task traffic prediction, which directly impacts the efficiency of task offloading strategies. Inspired by the recent success of large models, we propose a novel Lightweight Large Model with Temporal-Spatial Attention (L-LMTSA), designed to maximize task traffic prediction accuracy and minimize the overall computation latency. Specifically, L-LMTSA integrates a Temporal Attention Model (TAM) and a Spatial Attention Model (SAM) to capture multi-dimensional information, such as task flow location and temporal characteristics. In addition, we develop a pre-trained lightweight GPT-2 model by employing a one-dimensional convolutional input encoding layer and fine-tuning its parameters for predictive accuracy. Additionally, we optimize the UAV's trajectory and task offloading strategy using a three-dimensional artificial potential field informed by the predicted task traffic. Extensive experiments demonstrate that L-LMTSA not only improves training speed and prediction accuracy but also significantly reduces computational latency, resulting in more efficient UAV-assisted MEC experiences for consumers. © 2024 IEEE.

关键词： Unmanned aerial vehicles (UAV)

来源：评论

学校读者我要写书评

暂无评论

Database Scaling on Kubernetes 3

Database Scaling on Kubernetes

引用

3rd International Conference on Advancements in Computing, ICAC 2021

作者： Perera, H.C.S. De Silva, T.S.D. Wasala, W.M.D.C. Rajapakshe, R.M.P.R.L. Kodagoda, N. Samaratunge, Udara Srimath S. Jayanandana, H.H.N.C. Sri Lanka Institute of Information Technology Faculty of Computing Department of Computer Science and Software Engineering Malabe10115 Sri Lanka

ISBN: (纸本)9781665408622

Kubernetes is a hot topic in the field of software engineering and Distributed Computing. When compared to previous methods, the principle underlying Kubernetes, which is containerization, has altered how applications are created and delivered. However, when considering the state, particularly the databases, with Kubernetes, there is a scalability and data synchronization barrier. The most frequently used approach is to host the database outside of Kubernetes and maintain connectivity with the cluster. Kubernetes inherent capabilities are sufficient for hosting databases. But that requires high domain knowledge to do the configurations and maintain the databases on Kubernetes. The purpose of this research is to fulfil that gap by introducing a solution for managing highly available databases on Kubernetes. The solution is limited to managing PostgreSQL databases on Kubernetes using auto-scaling. A novel algorithm is proposed for auto-scaling, as previous algorithms do not take database requests into account when determining the scaling need. The drawbacks of data synchronization and auto-scaling will be solved in this research, and the end user will be able to access the service without interruption for the majority of the time, as the final solution makes the database cluster highly available for the service layer. © 2021 IEEE.

关键词： Database systems

来源：评论

学校读者我要写书评

暂无评论

A Fast CU Partition Decision Strategy for AVS3 Intra Coding

A Fast CU Partition Decision Strategy for AVS3 Intra Coding

引用

2022 IEEE International Conference on Consumer Electronics, ICCE 2022

作者： Chen, Changxin Luo, Xinjie Yan, Yunyao Xiang, Guoqing Zhang, Peng Huang, Xiaofeng Yan, Wei Peking University School of Software Microelectronics Beijing China Advanced Institute of Information Technology Peking University Beijing China School of Electronic and Computer Engineering Peking University Shenzhen China School of Communication Engineering Hangzhou Dianzi University Hangzhou China

ISBN: (纸本)9781665441544

Third generation audio and video coding standard (AVS3) is a latest video coding standard developed by China. AVS3 allows flexible coding unit (CU) partition by applying quad-tree (QT), binary tree (BT), and extended quad-tree (EQT) partitioning structures. The increased block partition flexibility brings the significant coding complexity cost. This paper proposes a fast CU partition decision method based on the gradient and the variance for AVS3 intra coding. Experimental results show that the proposed fast algorithm can achieve a good balance between computational complexity and coding performance. Compared with AVS3 software reference HPM4.0, the coding time is reduced by 43%, while the average BD-Rate is only increased by 0.9% in AI (All-Intra) configuration. © 2022 IEEE.

关键词： Binary trees

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：