检索结果-内蒙古大学图书馆

Tackling Class Imbalance and Client Heterogeneity for Split Federated Learning in Wireless Networks

IEEE Transactions on Wireless Communications 2025年第6期24卷 4920-4936页

作者： Xie, Chunfeng Chen, Zhixiong Yi, Wenqiang Shin, Hyundong Nallanathan, Arumugam Queen Mary University of London School of Electronic Engineering and Computer Science LondonE1 4NS United Kingdom University of Essex School of Computer Science and Electronic Engineering ColchesterCO4 3SQ United Kingdom Kyung Hee University Department of Electronics and Information Convergence Engineering Gyeonggido Yongin-si17104 Korea Republic of Kyung Hee University Department of Electronic Engineering Gyeonggi-do Yongin-si17104 Korea Republic of

As the complexity of deep neural networks escalates, traditional federated learning (FL) frameworks increasingly struggle since the training overhead of the full model is costly for resource-limited clients. In addition, the class imbalance among local datasets and client heterogeneity may lead to significant deterioration in learning performance. To address these challenges, we first propose a novel wireless split federated learning (SFL) framework to enhance learning efficiency and performance in resource-constrained networks, which adaptively splits the global model between the clients and server to alleviate the computation burden for clients. Then, we theoretically analyze how the client sampling and wireless network parameters impact on the convergence bound. Based on the analysis, we identify the extent of class imbalance that significantly impacts learning performance. Inspired by this, we formulate an optimization problem to strike a balance between latency and performance by jointly optimizing the client selection, model splitting, and bandwidth allocation policies. To solve this problem, we introduce a latency and class imbalance-aware double greedy algorithm to obtain client scheduling policy. Additionally, bisection-enabled optimal bandwidth allocation and model splitting algorithms are developed to adaptively determine bandwidth allocation and model splitting policies, respectively. Extensive experimental results demonstrate that our approach significantly reduces latency and enhances learning performance. © 2025 IEEE. All rights reserved.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

Very fast algorithms for implied barriers and moving-barrier options pricing

引用

Mathematics and computers in Simulation 2023年 205卷 251-271页

作者： Lu, Yu-Ming Lyuu, Yuh-Dauh Department of Computer Science & Information Engineering National Taiwan University No.30 Sec.3 Xinhai Road Taipei106032 Taiwan Department of Finance and Department of Computer Science & Information Engineering National Taiwan University No. 1 Sec. 4 Roosevelt Road Taipei10617 Taiwan

Two closely related O(nlogn)-time tree algorithms under the Black–Scholes model are presented, where n denotes the tree's number of time steps. The first finds the implied step barrier that matches the barrier-hitting probabilities exactly. In the constant-barrier case, the implied barrier is surprisingly accurate even for small ns;indeed, n=1 gives good results in typical situations. The second prices options with a time-dependent barrier (i.e., moving-barrier options). In practice, both algorithms are one to three orders faster than the standard algorithms even when n is moderate. As a consequence, large portfolios or datasets can finally be studied in a timely manner. Both algorithms can be easily tailored to handle barriers that are continuously monitored, discretely monitored, a mixture of both, or even when the model parameters are all time varying. © 2022 International Association for Mathematics and computers in Simulation (IMACS)

关键词： Fast Fourier transforms

来源：评论

学校读者我要写书评

暂无评论

Intelligent Proximate Analysis of Coal Based on Near-Infrared Spectroscopy and Multioutput Deep Learning

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2024年第3期5卷 1398-1410页

作者： Zou, Liang Qiao, Jiahui Yu, Xinhui Chen, Xun Lei, Meng School of Information and Control Engineering China University of Mining and Technology Xuzhou221116 China Department of Electrical and Computer Engineering University of British Columbia VancouverBCV6T 1Z4 Canada Department of Electronic Engineering and Information Science University of Science and Technology of China Hefei230027 China

Proximate analysis of coal indicates the moisture, ash, volatile content, and calorific value, which has been widely utilized as the basis for coal characterization. It involves heating the coal under various conditions until a constant weight is obtained. Although it is a relatively simple process that does not require expensive analytical equipment, determining these characteristics is time consuming. An alternative way for proximate analysis is spectral analysis in combination with various machine learning methods. However, most previous works analyze individual characteristics and fail to explore the relationship among them. In this study, we propose a method for proximate analysis based on near-infrared spectroscopy and a multioutput attention Unet (MOA-Unet), which can predict multiple characteristics simultaneously. First, an attention-based Unet is designed as the shared feature extraction subnetwork, including an encoder, a decoder, convolutional block attention modules, and multiscale feature fusion modules, which can improve the representation power of the U-shape network through aggregating features of shallower layers and concatenating features of deeper layers. Second, four individual subnetworks with fully connected layers, designed for four outputs, are utilized for regressing those four characteristics. We employ the gradient normalization algorithm to alleviate the gradient magnitude masking effect caused by training imbalance among different tasks. The proposedMOA-Unet is compared with classical chemometric methods on 670 coal samples from on-site *** experimental results demonstrate that the proposedmodel achieves state-of-the-art performance with correlation coefficients of 0.9015, 0.9538, 0.8986, and 0.8884, corresponding to moisture, ash, volatile content, and calorific value, respectively. Impact Statement-The proximate analysis of coal has been widely utilized as the basis for determining the rank of coal which is in connection with coa

关键词： Feature extraction

来源：评论

学校读者我要写书评

暂无评论

Optimizing Multidocument Summarization by Blending Reinforcement Learning Policies

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2023年第3期4卷 416-427页

作者： Su, DiJia Su, Difei Mulvey, John M. Poor, H.Vincent Princeton University Department of Electrical and Computer Engineering PrincetonNJ08540 United States University of British Columbia Department of Computer Science VancouverBCV6Z1T8 Canada Princeton University Department of Operation Research and Financial Engineering PrincetonNJ08540 United States

We consider extractive summarization within a cluster of related texts (multidocument summarization). Unlike single-document summarization, redundancy is particularly important because sentences across related documents might convey overlapping information. Thus, sentence extraction in such a setting is difficult because one will need to determine which pieces of information are relevant while avoiding unnecessary repetitiveness. To solve this difficult problem, we propose a novel reinforcement learning-based method Policy Blending with maximal marginal relevance and Reinforcement Learning (PoBRL) for solving multidocument summarization. PoBRL jointly optimizes over the following objectives necessary for a high-quality summary: importance, relevance, and length. Our strategy decouples this multiobjective optimization into different subproblems that can be solved individually by reinforcement learning. Utilizing PoBRL, we then blend each learned policies to produce a summary that is a concise and a complete representation of the original input. Our empirical analysis shows high performance on several multidocument datasets. Human evaluation also shows that our method produces high-quality output. © 2020 IEEE.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Network-Based Rail Running Band Anomaly Recognition via Recurrent Attention Graphs

引用

IEEE Transactions on Automation science and engineering 2025年 22卷 15381-15395页

作者： Ni, Xuefeng Fieguth, Paul W. Ma, Ziji Shi, Bo Liu, Hongli Hunan University College of Electrical and Information Engineering Changsha410012 China University of Waterloo Department of Systems Design Engineering WaterlooONN2L 3G1 Canada The Hong Kong University of Science and Technology Department of Computer Science and Engineering Sai Kung Hong Kong

Anomaly detection for rail running bands, the pattern of wheel-rail contact area, is crucial to analyze composite rail irregularities. This paper presents an all-weather vision-based solution for running-band inspection. However, two major algorithmic challenges restrict inspection effectiveness: 1) the identification of high-quality features for diversified and imbalanced data subject to noise and outliers, and 2) inferring implicit anomaly co-occurrence patterns. We regard overall running-band anomaly detection as a multi-label classification problem and develop a novel deep multi-anomaly recognition network via recurrent attention graphs (RAGRN). © 2004-2012 IEEE.

关键词： Network theory (graphs)

来源：评论

学校读者我要写书评

暂无评论

Toward a Neural Network-Based Approach for Improved Atmospheric Infrasound Localization

引用

IEEE Access 2025年 13卷 73745-73754页

作者： Wu, Addison J. Joshi, Arnav Hickey, Jean-Pierre Princeton University Department of Computer Science PrincetonNJ08544 United States University of Waterloo Department of Mechanical and Mechatronics Engineering WaterlooONN2L 3G1 Canada Massachusetts Institute of Technology Department of Mechanical Engineering CambridgeMA02139 United States

Accurate infrasound localization represents a challenging problem in acoustics that yields many important applications. Current methods for distance and altitudinal localization have only been evaluated in limited short-range scenarios and fail to take full advantage of real-time atmospheric surroundings to maximize accuracy. We develop a neural network architecture that uses both the local acoustic pressure and atmospheric data to produce accurate localization predictions including the lateral distance and altitude of an infrasonic source, using only a singular-sensor arrangement for a more flexible and practical implementation that can be transferred to real-time aerial detection. To improve generalizability, we train our neural network in a variety of atmospheric conditions and conduct validation testing on samples generated from designated hold-out atmospheric profiles. We generated a comprehensive infrasonic raypath dataset using an experimentally validated infrasound raytracing tool, a technique inspired by other fields where controlled data is difficult to obtain in real life. A preliminary comparison with existing infrasound localization methods in similar conditions demonstrated that our neural network-based approach approximately halved both median distance and altitudal localization errors. These promising results highlight the potential of our neural network-based approach to significantly enhance infrasound localization accuracy across diverse atmospheric conditions. Future steps would involve more thorough validation and benchmarking of the model in real-world scenarios to detect any trends in localization error that were not captured in this work. © 2013 IEEE.

关键词： Aircraft detection

来源：评论

学校读者我要写书评

暂无评论

A machine learning-guided design and manufacturing of wearable nanofibrous acoustic energy harvesters

引用

Nano Research 2024年第10期17卷 9181-9192页

作者： Negar Hosseinzadeh Kouchehbaghi Maryam Yousefzadeh Aliakbar Gharehaghaji Safoora Khosravi Danial Khorsandi Reihaneh Haghniaz Ke Cao Mehmet R.Dokmeci Mohammad Rostami Ali Khademhosseini Yangzhi Zhu Terasaki Institute for Biomedical Innovation Los AngelesCA 91367USA Department of Textile Engineering Amirkabir University of Technology(Tehran Polytechnic)Hafez Avenue*** TehranIran Department of Electrical and Computer Engineering University of British ColumbiaVancouverBC V6T1Z4Canada Chemical Sciences Division Oak Ridge National LaboratoryOak RidgeTN 37831USA Department of Computer Science University of Southern CaliforniaLos AngelesCA 90007USA

Nanofibrous acoustic energy harvesters(NAEHs)have emerged as promising wearable platforms for efficient noise-to-electricity conversion in distributed power energy systems and wearable sound amplifiers for assistive listening ***,their reallife efficacy is hampered by low power output,particularly in the low-frequency range(<1 kHz).This study introduces a novel approach to enhance the performance of NAEHs by applying machine learning(ML)techniques to guide the synthesis of electrospun polyvinylidene fluoride(PVDF)/polyurethane(PU)nanofibers,optimizing their application in wearable *** use a feed-forward neural network along with solving an optimization problem to find the optimal input values of the electrospinning(applied voltage,nozzle-collector distance,electrospinning time,and drum rotation speed)to generate maximum output performance(acoustic-to-electricity conversion efficiency).We first prepared a dataset to train the network to predict the output power given the input variables with high *** introducing the neural network,we fix the network and then solve an optimization problem using a genetic algorithm to search for the input values that lead to the maximum energy harvesting *** ML-guided wearable PVDF/PU NAEH platform can deliver a maximal acoustoelectric power density output of 829μW/cm^(3) within the surrounding noise *** addition,our system can function stably in a broad frequency(0.1-2 kHz)with a high energy conversion efficiency of 66%.Sound recognition analysis reveals a robust correlation exceeding 0.85 among lexically akin terms with varying sound intensities,contrasting with a diminished correlation below 0.27 for words with disparate semantic ***,this work provides a previously unexplored route to utilize ML in advancing wearable NAEHs with excellent practicability.

关键词： wearable electronics acoustic energy harvester machine learning piezoelectric nanogenerator electrospun nanofiber

来源：评论

学校读者我要写书评

暂无评论

A Comprehensive Model Incorporating Multiple Spatial Relations in 3D Space

引用

Recent Advances in computer science and Communications 2023年第8期16卷 65-77页

作者： Liu, Weiguang Li, Mengmeng Zhao, Yuanyuan Gao, Jixun Wang, Miao Fang, Zhenxi Department of Computer Science Zhongyuan University of Technology No.41 Zhongyuan Road Henan Zhongyuan District China Propaganda Department Zhengzhou University of Technology Henan No.108 Yingcai Road Huiji District China Department of Computer Science Henan University of Engineering No.1 Xianghe Road Longhu Henan Zhengzhou China

Aims: A Comprehensive Model Incorporating Multiple Spatial Relations in 3D Space. Background: At present, the research on two-dimensional spatial relation expression and inference models is relatively mature, but these models cannot be used to deal with three-dimensional spatial orientation relations. With the application of spatial orientation relations, three-dimensional spatial orientation relations are involved in many fields such as urban architectural design, robotics, image processing, etc. Two-dimensional spatial orientation relations models cannot satisfy the needs of three-dimensional spatial applications, so there is an urgent need to research three-dimensional spatial object orientation relations expression and inference models. Objective: This work aims to provide a comprehensive model incorporating multiple spatial relations in 3D space. The joint representation of direction, distance, and topological relations describes more complex spatial position relations among objects. Method: Based on this comprehensive model, the computational properties of interval algebra are used to combine the directional and topological relations. Result: The study lays a good foundation for the formal representation and reasoning of spatial relations between regions, enhances the analyzability of spatial relations between objects, and improves the accuracy of spatial analysis. Conclusion: The main novel contribution of this paper is that we propose a comprehensive orientation relation model, called 3D-TRD, which considers three spatial orientation relations sim-ultaneously. The paper gives examples to represent the position relations of two spatial objects by comparing the RCC8 model, the 3DR46 model, and the comprehensive model to highlight the advantages of our proposed model. Based on the model, the bidirectional mapping association method is also used to represent the location of the spatial objects. The first advantage of the 3D-TRD model is that it represents spatia

关键词： Three dimensional computer graphics

来源：评论

学校读者我要写书评

暂无评论

Efficient Prompting for LLM-Based Generative Internet of Things

引用

IEEE Internet of Things Journal 2025年第1期12卷 778-791页

作者： Xiao, Bin Kantarci, Burak Kang, Jiawen Niyato, Dusit Guizani, Mohsen University of Ottawa School of Electrical Engineering and Computer Science OttawaONK1N 6N5 Canada Guangdong University of Technology School of Automation Guangzhou510006 China Nanyang Technological University School of Computer Science and Engineering Jurong West Singapore Singapore Mohamed bin Zayed University of Artificial Intelligence Machine Learning Department Abu Dhabi United Arab Emirates

Large language models (LLMs) have demonstrated remarkable capacities on various tasks, and integrating the capacities of LLMs into the Internet of Things (IoT) applications has drawn much research attention recently. Due to security concerns, many institutions avoid accessing state-of-the-art commercial LLM services, requiring the deployment and utilization of open-source LLMs in a local network setting. However, open-source LLMs usually have more limitations regarding their performance, such as their arithmetic calculation and reasoning capacities, and practical systems of applying LLMs to IoT have yet to be well-explored. Therefore, we propose an LLM-based Generative IoT (GIoT) system deployed in the local network setting in this study. To alleviate the limitations of LLMs and provide service with competitive performance, we apply prompt engineering methods to enhance the capacities of the open-source LLMs, design a Prompt Management Module and a Postprocessing Module to manage the tailored prompts for different tasks and process the results generated by the LLMs. To demonstrate the effectiveness of the proposed system, we discuss a challenging table question answering (Table-QA) task as a case study of the proposed system, as tabular data is usually more challenging than plaintext because of their complex structures, heterogeneous data types and sometimes huge sizes. We conduct comprehensive experiments on the two popular Table-QA data sets, and the results show that our proposal can achieve competitive performance compared with state-of-the-art LLMs, demonstrating that the proposed LLM-based GIoT system can provide competitive performance with tailored prompting methods and is easily extensible to new tasks without training. © 2014 IEEE.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Identification and Visualization of Spatial and Temporal Trends in Textile Industry

引用

computers, Materials & Continua 2023年第2期74卷 4165-4181页

作者： Umair Yousaf Muhammad Asif Shahbaz Ahmed Noman Tahir Azeem Irshad Akber Abid Gardezi Muhammad Shafiq Jin-Ghoo Choi Habib Hamam Department of Computer Science National Textile UniversityFaisalabad38000Pakistan Department of Computer Science and Software Engineering International Islamic UniversityIslamabad44000Pakistan Department of Computer Science COMSATS University IslamabadIslamabad45550Pakistan Department of Information and Communication Engineering Yeungnam UniversityGyeongsan38541Korea Faculty of Engineering Uni de MonctonMonctonNB E1A3E9Canada International Institute of Technology and Management Commune d’AkandaBPLibreville1989Gabon School of Electrical Engineering Department of Electrical and Electronic Eng.ScienceUniversity of JohannesburgJohannesburg2006South Africa Spectrum of Knowledge Production&Skills Development Sfax3027Tunisia

The research volume increases at the study rate,causing massive text *** to these enormous text corpora,we are drowning in data and starving for ***,recent research employed different text mining approaches to extract information from this text *** proposed approaches extract meaningful and precise phrases that effectively describe the text’s *** extracted phrases are commonly termed ***,these key phrases are employed to determine the different fields of study ***,these key phrases can also be used to determine the spatiotemporal trends in the various research *** this research,the progress of a research field can be better revealed through spatiotemporal bibliographic trend ***,an effective spatiotemporal trend extraction mechanism is required to disclose textile research trends of particular regions during a specific *** study collected a diversified dataset of textile research from 2011–2019 and different countries to determine the research *** data was collected from various open access ***,this research determined the spatiotemporal trends using quality *** research also focused on finding the research collaboration of different countries in a particular research *** research collaborations of other countries’researchers show the impact on import and export of those *** visualization approach is also incorporated to understand the results better.

关键词： Text mining spatiotemporal trend research collaboration

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：