检索结果-内蒙古大学图书馆

A survey on cross-user federated recommendation

science China(Information sciences) 2025年第4期68卷 7-32页

作者： Enyue YANG Yudi XIONG Wei YUAN Weike PAN Qiang YANG Zhong MING College of Computer Science and Software Engineering Shenzhen University School of Electrical Engineering and Computer Science The University of Queensland WeBank AI Lab WeBank Department of Computer Science and Engineering Hong Kong University of Science and Technology College of Big Data and Internet Shenzhen Technology University Guangdong Laboratory of Artificial Intelligence and Digital Economy (SZ)

Recommender systems are effective in mitigating information overload, yet the centralized storage of user data raises significant privacy concerns. Cross-user federated recommendation(CUFR) provides a promising distributed paradigm to address these concerns by enabling privacy-preserving recommendations directly on user devices. In this survey, we review and categorize current progress in CUFR, focusing on four key aspects: privacy, security, accuracy, and efficiency. Firstly,we conduct an in-depth privacy analysis, discuss various cases of privacy leakage, and then review recent methods for privacy protection. Secondly, we analyze security concerns and review recent methods for untargeted and targeted *** untargeted attack methods, we categorize them into data poisoning attack methods and parameter poisoning attack methods. For targeted attack methods, we categorize them into user-based methods and item-based methods. Thirdly,we provide an overview of the federated variants of some representative methods, and then review the recent methods for improving accuracy from two categories: data heterogeneity and high-order information. Fourthly, we review recent methods for improving training efficiency from two categories: client sampling and model compression. Finally, we conclude this survey and explore some potential future research topics in CUFR.

关键词： cross-user federated recommendation federated recommendation federated learning recommender systems user privacy

来源：评论

学校读者我要写书评

暂无评论

UniArray: Unified Spectral-Spatial Modeling for Array-Geometry-Agnostic Speech Separation

引用

IEEE Signal Processing Letters 2025年 32卷 2164-2168页

作者： Chen, Weiguang Zhang, Junjie Yang, Jielong Chng, Eng Siong Zhong, Xionghu Hunan University College of Computer Science and Electronic Engineering Changsha410082 China Jiangnan University School of Internet of Things Engineering Wuxi214122 China Nanyang Technological University School of Computer Science and Engineering 639798 Singapore

Array-geometry-agnostic speech separation (AGA-SS) aims to develop an effective separation method regardless of the microphone array geometry. Conventional methods rely on permutation-free operations, such as summation or attention mechanisms, to capture spatial information. However, these approaches often incur high computational costs or disrupt the effective use of spatial information during intra- and inter-channel interactions, leading to suboptimal performance. To address these issues, we propose UniArray, a novel approach that abandons the conventional interleaving manner. UniArray consists of three key components: a virtual microphone estimation (VME) module, a feature extraction and fusion module, and a hierarchical dual-path separator. The VME ensures robust performance across arrays with varying channel numbers. The feature extraction and fusion module leverages a spectral feature extraction module and a spatial dictionary learning (SDL) module to extract and fuse frequency-bin-level features, allowing the separator to focus on using the fused features. The hierarchical dual-path separator models feature dependencies along the time and frequency axes while maintaining computational efficiency. Experimental results show that UniArray outperforms state-of-the-art methods in SI-SDRi, WB-PESQ, NB-PESQ, and STOI across both seen and unseen array geometries. © 1994-2012 IEEE.

关键词： Computational efficiency

来源：评论

学校读者我要写书评

暂无评论

Research on resource allocation methods for traditional Chinese medicine services based on deep reinforcement learning

引用

Neural Computing and Applications 2025年第3期37卷 1601-1616页

作者： Ma, Yuntao Fang, Xiaolin Qi, Jin Sun, Yanfei School of Computer Science and Engineering Southeast University Jiangsu Nanjing China School of Internet of Things Nanjing University of Posts and Telecommunications Jiangsu Nanjing China

Chinese medicine resources are the crystallization of traditional Chinese culture, and more and more people are choosing Chinese medicine services for their health. To address the problems of pluralistic heterogeneity, waste of service resources, and lagging demand response in the resource allocation model for traditional Chinese medicine (TCM) services, a deep reinforcement learning-based resource allocation method for TCM services is proposed. To address the fragmentation of TCM service resources, this paper presents a TCM service resource association method based on improved spectral clustering and establishes a good resource-demand matching model. For the problem of TCM service resource allocation after resource association, we establish a TCM service resource allocation model and collaboratively solve the TCM service resource allocation problem via the deep reinforcement learning method. The results show that the proposed solution can accelerate the demand response of TCM service resources, effectively reduce the cost of TCM services for patients, improve the quality of TCM services, and satisfy the demand for TCM services for patients. © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2024.

关键词： Deep reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Target Detection-Oriented RGCN Inference Enhancement Method

引用

computers, Materials & Continua 2025年第4期83卷 1219-1237页

作者： Lijuan Zhang Xiaoyu Wang Songtao Zhang Yutong Jiang Dongming Li Weichen Sun School of Internet of Things Engineering Wuxi UniversityWuxi214105China College of Computer Science and Engineering Changchun University of TechnologyChangchun130012China Computer Science and Technology at Guohao Academy Tongji UniversityShanghai200092China China North Vehicle Research Institute Beijing100072China

In this paper,a reasoning enhancement method based on RGCN(Relational Graph Convolutional Network)is proposed to improve the detection capability of UAV(Unmanned Aerial Vehicle)on fast-moving military targets in urban battlefield *** combining military images with the publicly available VisDrone2019 dataset,a new dataset called VisMilitary was built and multiple YOLO(You Only Look Once)models were tested on *** to the low confidence problem caused by fuzzy targets,the performance of traditional YOLO models on real battlefield images decreases ***,we propose an improved RGCN inference model,which improves the performance of the model in complex environments by optimizing the data processing and graph network *** results show that the proposed method achieves an improvement of 0.4%to 1.7%on mAP@0.50,which proves the effectiveness of the model in military target *** research of this paper provides a new technical path for UAV target detection in urban battlefield,and provides important enlightenment for the application of deep learning in military field.

关键词： RGCN target detection urban battlefield YOLO visual reasoning

来源：评论

学校读者我要写书评

暂无评论

AI-Driven Energy Optimization in UAV-Assisted Routing for Enhanced Wireless Sensor Networks Performance

引用

computers, Materials & Continua 2024年第9期80卷 4085-4110页

作者： Syed Kamran Haider Abbas Ahmed Noman Mujeeb Khan Ali Nauman Sung Won Kim College of Internet of Things(IoT)Engineering Hohai UniversityChangzhou213001China Department of Electrical and Electronics Engineering Beaconhouse International CollegeIslamabad44000Pakistan School of Computer Science and Engineering Yeungnam UniversityGyeongsan38428Republic of Korea

In recent advancements within wireless sensor networks(WSN),the deployment of unmanned aerial vehicles(UAVs)has emerged as a groundbreaking strategy for enhancing routing efficiency and overall network *** research introduces a sophisticated framework,driven by computational intelligence,that merges clustering techniques with UAV mobility to refine routing strategies in *** proposed approach divides the sensor field into distinct sectors and implements a novel weighting system for the selection of cluster heads(CHs).This system is primarily aimed at reducing energy consumption through meticulously planned routing and path *** a greedy algorithm for inter-cluster dialogue,our framework orchestrates CHs into an efficient communication *** comparative analysis,the proposed model demonstrates a marked improvement over traditional methods such as the cluster chain mobile agent routing(CCMAR)and the energy-efficient cluster-based dynamic algorithms(ECCRA).Specifically,it showcases an impressive 15%increase in energy conservation and a 20%reduction in data transmission time,highlighting its advanced ***,this paper investigates the impact of various network parameters on the efficiency and robustness of the WSN,emphasizing the vital role of sophisticated computational strategies in optimizing network operations.

关键词： UAV trajectory clustering next-generation wireless sensor network(NGWSN) energy efficiency mobile sink

来源：评论

学校读者我要写书评

暂无评论

ISAC-Fi: Enabling Full-Fledged Monostatic Sensing over Wi-Fi Communication

IEEE Journal of Selected Areas in Sensors

引用

IEEE Journal of Selected Areas in Sensors 2024年 1卷 139-153页

作者： Chen, Zhe Hu, Chao Zheng, Tianyue Cao, Hangcheng Yang, Yanbing Chu, Yen Jiang, Hongbo Luo, Jun Fudan University Institute of Space Internet School of Computer Science Shanghai200437 China Sichuan University College of Computer Science Chengdu610017 China Nanyang Technological University School of Computer Science and Engineering Singapore639798 Singapore Hunan University College of Computer Science and Electronics Engineering Changsha410012 China University of Electronic Science and Technology of China Chengdu610054 China

Although Wi-Fi communications have been exploited for sensing purpose for over a decade, the bistatic or multistatic nature of Wi-Fi still poses multiple challenges, hampering real-life deployment of integrated sensing and communication (ISAC) within the Wi-Fi framework. In this article, we aim to redesign Wi-Fi so that monostatic sensing (mimicking radar) can be achieved over the multistatic communication infrastructure. Specifically, we propose, design, and implement ISAC-Fi as an ISAC-ready Wi-Fi prototype. We first present a novel self-interference cancellation scheme, in order to extract reflected (radio frequency) signals for sensing purpose in the face of transmissions. We then subtly revise the existing Wi-Fi framework so as to seamlessly operate monostatic sensing under the Wi-Fi communication standard. Finally, we offer two ISAC-Fi designs: 1) a Universal Software Radio Peripheral (USRP)-based design emulates a totally redesigned ISAC-Fi device and 2) another plug-and-play design allows for backward compatibility by attaching an extra module to an arbitrary Wi-Fi device. We perform extensive experiments to validate the efficacy of ISAC-Fi and also to demonstrate its superiority over existing Wi-Fi sensing proposals. © 2024 IEEE.

关键词： Wi-Fi

来源：评论

学校读者我要写书评

暂无评论

nPGSAO: A Hybrid Particle Swarm Optimization and Genetic Algorithm with Niching Technology for Edge Server Placement

引用

IEEE internet of Things Journal 2025年第12期12卷 19370-19383页

作者： Wang, Bo Zhang, Zhifeng Song, Ying Chen, Ming Liu, Dongqing Software Engineering College Zhengzhou University of Light Industry Zhengzhou China Computer School Beijing Information Science and Technology University Beijing China School of Mathematics and Computer Science Tongling University Tongling China National Engineering Laboratory for Internet Medical Systems and Applications the First Affiliated Hospital of Zhengzhou University Zhengzhou China

In this paper, we focus on the edge server placement problem (ESPP). ESPP is to decide positions (edge stations) where purchased edge servers (ESs) are placed when a service provider builds or upgrades its edge computing solution. The decision of ESPP largely determines the overall service quality and resource efficiency. We aim at providing ESPP solutions for optimizing the service quality in terms of the system response time, given a set of purchased ESs. To achieve this aim, we first deduce the optimal request dispatch for each edge station based on queue theory and Binomial theorem. Then, based on the deduced result, we formulate ESPP into a discrete nonlinear constrained optimization model. Because ESPP is NP-hard and generally multi-modality, we designed a hybrid metaheuristic algorithm to combine advantages of particle swarm optimization (PSO) and genetic algorithm (GA) as well as the niching technology, which is named nPGSAO. The basic idea of nPGSAO is grouping similar individuals into a niche and finding the niche best. Then, nPGSAO evolves each individual by integrating the swarm intelligence of PSO into the evolutionary process of GA based on its personal and niching best. To evaluate the performance of nPGSAO, we conducted extensive simulated experiments. The results show that nPGSAO achieves 5.78%–40.4% better response time with varied ESPP scale, compared to 11 up-to-date algorithms, and has a linear time complexity with the ESPP scale. These confirm the superior performance and scalability of nPGSAO. © 2014 IEEE.

关键词： Genetic algorithms

来源：评论

学校读者我要写书评

暂无评论

High Order Time Shift Keying Modulation for Ambient Backscatter Communications

引用

IEEE Transactions on Communications 2024年第6期73卷 3792-3803页

作者： Chen, Jixiang Guan, Quansheng Rong, Yue Li, Dong Chen, Weiqi Yu, Hua South China University Of Technology School Of Electronic And Information Engineering Guangzhou510640 China Curtin University School Of Electrical Engineering Computing And Mathematical Sciences BentleyWA6102 Australia Macau University Of Science And Technology School Of Computer Science And Engineering 999078 China Guangdong University Of Finance School Of Internet Finance And Information Engineering Guangzhou China

Ambient backscatter communication (AmBC) is a newly cutting-edge technology for the internet of Things, which utilizes the ambient radio frequency signal as the carrier to transmit information. Existing works focus on the simple on-off keying modulation which has low channel utilization. However, it is not desirable to develop the high-order modulation in the power domain due to the weak strength of the backscattered signal. In this paper, we extend the high-order modulation in the time domain instead, i.e., high-order time shift keying (TSK). Since the channel coherent time is unknown at the receiver and the tag, the detection methods with training symbols will have a huge performance degradation if the channels are changed and training symbols become outdated. To overcome this challenge, we further propose the transition-aided TSK (TA-TSK) modulation and the frequency-shifting TSK (FS-TSK) modulation, which do not need to send training symbols at the tag. These two methods can work well even if the channel coherent time is as short as one time slot. Meanwhile, the detection methods for TSK are developed and the corresponding closed-form bit-error-rate (BER) expressions are obtained. Simulation results show that a high modulation order is more suitable for the M-ary phase shift keying source than the complex Gaussian source. The high order TSK can provide at least 2 dB signal-to-noise ratio (SNR) gain at the same BER compared to the on-off keying. © 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

E-Waste Management using Deep Learning 6

E-Waste Management using Deep Learning

引用

6th International Conference on Mobile Computing and Sustainable Informatics, ICMCSI 2025

作者： Kumar, Boddu Manoj Sujitha, Kaduru Abinaya, K. School of Computer Science Engineering with Blockchain Technology Sathyabama Institute of Science and Technology Tamil Nadu Chennai600119 India School of Computer Science Engineering with Internet of Things Sathyabama Institute of Science and Technology Tamil Nadu Chennai600119 India School of Computer Science and Engineering Sathyabama Institute of Science and Technology Tamil Nadu Chennai600119 India

ISBN: (纸本)9798331522667

Using the YOLO v8 model and deep learning approaches, this study explores the field of e-waste management and provides effective item detection. Our research attempts to increase the accuracy and scalability of e-waste detection, simplifying its incorporation into current waste management systems. Through utilizing deep learning, our goal is to transform the way that electronic waste is categorized, from tiny parts to larger appliances, allowing for more efficient as well as environmentally friendly management techniques. When it comes to aiming for higher performance, privacy comes first. Methods of preservation guarantee the accuracy of sensitive data correlated with pictures of e-waste. By means of execution utilizing data anonymization and encryption methods, we ensure compliance with legal requirements while safeguarding against potential privacy threats and ethical issues. The YOLO v8 model achieved an average precision (Map) of 97.1% in detecting e-waste items, demonstrating its effectiveness and scalability for real-world applications. © 2025 IEEE.

关键词： Differential privacy

来源：评论

学校读者我要写书评

暂无评论

Covid-19 Classification using Fine-tuned EfficientNet Architecture 9

Covid-19 Classification using Fine-tuned EfficientNet Archit...

引用

9th IEEE International Conference for Convergence in Technology, I2CT 2024

作者： Sruthi, Srigiri Emadaboina, Siddharth MacHavarapu, Pradyumna Singh, Rimjhim Padam Kanchan, Sneha Amrita Vishwa Vidyapeetham Amrita School of Computing Department of Computer Science And Engineering Bengaluru India Universiti Tunku Abdul Rahman Department of Internet Engineering And Computer Science Malaysia

ISBN: (纸本)9798350394474

This work utilizes an Efficient Net-based deep learning architecture, leveraging its efficiency and scalability for image classification tasks. We employ a Dataset of chest X-ray images, including COVID-19 cases as well as other relevant Pathology, to train and evaluate the model. The training process involves data augmentation, model optimization, and early stopping to ensure robust and generalizable performance. It specifies the image size, channels, and class count, and then proceeds to create a pre-trained model using the EfficientNetB0 architecture. The model is built upon the pre-trained EfficientNetB0 model, with additional layers including Layer Normalization, Dense, and Dropout layers for fine-tuning on the Covid-19 data. Subsequently, the model is rigorously evaluated using standard performance metrics, including precision, recall, F1-score, and accuracy, and has proven to outperform other state-of-art models by a 3% increase in F1-score value. © 2024 IEEE.

关键词： Deep learning

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：