检索结果-内蒙古大学图书馆

A Multi-Task Based clustering Personalized Federated Learning Method

Big Data Mining and Analytics 2024年第4期7卷 1017-1030页

作者： Ao Xiong Han Zhou Yu Song Dong Wang Xu Wei Da Li Bo Gao State Key Laboratory of Networking and Switching Technology Beijing University of Posts and TelecommunicationsBeijing 100876China State Grid Digital Technology Holding Co.Ltd. Beijing 100053China State Grid Jiangsu Electric Power Company Co.Ltd. Nanjing 210008China

Federated Learning(FL)is a framework for machine learning on a large-scale distributed dataset,enabling the training of a collaborative model across multiple parties while preserving the privacy of user ***,in cases where data are distributed in a non-independent and identically distributed(non-iid)manner,the convergence speed of the federated collaborative model and its prediction accuracy on client nodes can be significantly ***,personalized FL methods have emerged to further adapt to the data characteristics of different *** response to the data heterogeneity issue,this paper presents a multi-task clustering-based personalized federated learning algorithm,which is applied to the prediction of carbon emissions in different regions and *** algorithm partitions nodes with similar data distributions and aggregates local models within the same cluster to form cluster *** introduces the concept of multitask learning,dividing the lower layers of cluster models into expert *** expert layers of different cluster models are then weighted and aggregated for the acquisition of global ***,adaptive weight is applied to control the aggregation of expert layers,thereby achieving model personalization at the local *** experiments conducted on carbon emission prediction data demonstrate that the proposed algorithm performs better in various evaluation metrics compared with the Federated Averaging(FedAvg)algorithm and traditional clustering personalized federated learning *** also exhibits excellent experimental results and performance when dealing with different quantities of heterogeneous data distributions.

关键词： personalized federated learning data heterogeneity clustering algorithm multi-task learning adaptive weight

来源：评论

学校读者我要写书评

暂无评论

Digital Design Process and Part Family Division of Solid Wood Custom Cabinet Door Based on Multi-attribute Overlapping clustering Technology

引用

BIORESOURCES 2022年第3期17卷 5393-5419页

作者： Ren, Jie Xiong, Xianqing Nanjing Forestry Univ Coll Furnishings & Ind Design Nanjing 210037 Jiangsu Peoples R China Nanjing Forestry Univ Coll Furnishings & Ind Design Coinnovat Ctr Efficient Proc & Utilizat Forest Re Nanjing 210037 Jiangsu Peoples R China

The flexible production of solid wood custom furniture is needed to meet the individual needs of the market. Based on M Company, this study developed a digital design platform and its key technologies (data extraction and integration, part family division, coding technology, model base construction, information interaction, and connection) and proposed a multi-attribute overlapping clustering technology (MOCT) method for the division of a family of cabinet door parts. The results indicated that the platform and technologies based on M company provides a reference for the digital transformation of solid wood furniture. The part family division method based on MOCT displayed excellent grouping performance and adjustability, and it could be extensively applied to the group processing of solid wood furniture parts.

关键词： Multi-attribute decision-making clustering algorithm Solid wood furniture Digital design Cabinet door

来源：评论

学校读者我要写书评

暂无评论

clustering for smart cities in the internet of things: a review

引用

CLUSTER COMPUTING-THE JOURNAL OF NETWORKS SOFTWARE TOOLS AND APPLICATIONS 2022年第6期25卷 4097-4127页

作者： Hosseinzadeh, Mehdi Hemmati, Atefeh Rahmani, Amir Masoud Iran Univ Med Sci Psychosocial Hlth Res Inst Mental Hlth Res Ctr Tehran Iran Univ Human Dev Dept Comp Sci Sulaymaniyah Iraq Natl Yunlin Univ Sci & Technol Future Technol Res Ctr 123 Univ RdSect 3 Touliu 64002 Yunlin Taiwan

Nowadays, internet of things (IoT) applications, especially in smart cities, are fast developing. clustering is a promising solution for handling IoT issues such as energy efficiency, scalability, robustness, mobility, load balancing, and so on. The clustering method, which can be applied in IoT, groups sensor nodes into clusters with one node operating as the cluster head. This paper intends to determine the usage of clustering in IoT as a case study for smart cities. Furthermore, this study discusses clustering algorithms on IoT, open issues, and future challenges of clustering in the context of the smart city, and also existing research papers selected by the systematic literature review technique published between 2017 and 2021. Also, we provide a technical taxonomy for clustering categorization in IoT, which includes algorithm, architecture, and application. According to the statistical analysis of 51 chosen research articles in the domain of clustering in IoT, the results show that the number of clusters has a high percentage of 24%, the energy factor has 23%, the execution time factor has 18%, the accuracy has 14%, the delay has 9%, the lifetime has 6%, and throughput has 6%.

关键词： Internet of Things clustering algorithm Smart city Systematic literature review

来源：评论

学校读者我要写书评

暂无评论

Research on territorial spatial planning based on data mining and geographic information visualization

引用

NONLINEAR ENGINEERING - MODELING AND APPLICATION 2025年第1期14卷

作者： Wang, Yingqiao Planning Res & Informat Ctr Sanya Municipal Bur Nat Resources & Planning Hainan 572000 Peoples R China

At present, big data and mining technology has been more and more applied to urban planning practice. Based on data mining technology and geographic information visualization, this article puts forward a study on the optimization of territorial spatial planning. This article selects the clustering algorithm and decision tree in data mining technology, combines them organically, and puts forward spatial data mining. Cluster analysis is used to conduct spatial cluster analysis of comprehensive spatial data to explore whether there is spatial correlation and large-scale contiguous data in the data;if so, it can realize the purpose of national territorial space division. According to the analysis of the relevant data of land use planning and the characteristics of the business process, the overall design of the land use planning decision support system was carried out. The preliminary effects of land use planning clustering were generated through the usage of the K-central factor clustering algorithm, and the consequences with excessive classification accuracy have been chosen as coaching samples mixed with the baseline of time series. Finally, the K nearest neighbor classification algorithm primarily based on dynamic time warping is used to classify and perceive the city's practical place again. With the useful resource of point of interest data, the remaining city characteristic identification result is obtained. Finally, the simulation analysis shows that the area under curve (AUC) value of seven types of land use is 0.9376 for cultivated land, 0.85442 for forested land, 0.81747 for grassland, 0.8708 for water area, and 0.86672 for rural residential area. The AUC value of other construction land is 0.80346, and the AUC value of urban construction land is 0.9376. In the urban expansion planning project, multi-source data such as land use status data, population distribution data, and economic development data are integrated, and the spatial agglomeration pattern of la

关键词： data mining clustering algorithm decision tree time series national space planning

来源：评论

学校读者我要写书评

暂无评论

Estimation of water quality variables based on machine learning model and cluster analysis-based empirical model using multi-source remote sensing data in inland reservoirs, South China

引用

ENVIRONMENTAL POLLUTION 2024年 342卷 123104页

作者： Tian, Di Zhao, Xinfeng Gao, Lei Liang, Zuobing Yang, Zaizhi Zhang, Pengcheng Wu, Qirui Ren, Kun Li, Rui Yang, Chenchen Li, Shaoheng Wang, Meng He, Zhidong Zhang, Zebin Chen, Jianyao Sun Yat Sen Univ Sch Geog & Planning Guangzhou 510275 Peoples R China Zhuhai Ecol Environm Monitoring Stn Guangdong Prov Zhuhai 519070 Peoples R China Chinese Acad Sci Key Lab Vegetat Restorat & Management Degraded Eco South China Bot Garden Guangzhou 510650 Peoples R China Chinese Acad Geol Sci Inst Karst Geol Key Lab Karst Dynam Minist Nat Resources & Guangxi 50 Qixing Rd Guilin 541004 Guangxi Peoples R China

Reservoirs play important roles in the drinking water supply for urban residents, agricultural water provision, and the maintenance of ecosystem health. Satellite optical remote sensing of water quality variables in medium and micro-sized inland waters under oligotrophic and mesotrophic status is challenging in terms of the spatiotemporal resolution, weather conditions and frequent nutrient status changes in reservoirs, etc., especially when quantifying non-optically active components (non-OACs). This study was based on the surface reflectance products of unmanned aerial vehicle (UAV) multispectral images, Sentinel-2B Multispectral instrument (MSI) images and Landsat 7 Enhanced Thematic Mapper Plus (ETM+) by utilizing fuzzy C-means (FCM) clustering algorithm was combined with band combination (BC) model to construct the FCM-BC empirical model, and used mixed density network (MDN), extreme gradient boosting (XGBoost), deep neural network (DNN) and support vector regression (SVR) machine learning (ML) models to invert 12 kinds of optically active components (OACs) and non-OACs. Compared with the unclustered BC (UC) model, the mean coefficient of determination (MR) of the FCM-BC models was improved by at least 46.9%. MDN model showed best accuracy (R2 in the range of 0.60-0.98) and stability (R2 decreased by up to 13.2%). The accuracy of UAV was relatively higher in both empirical methods and machine learning methods. Additionally, the spatio-temporal distribution maps of four water quality variables were mapped based on the MDN model and UAV images, all platforms showed good consistency. An inversion strategy of water quality variables in various monitoring frequencies and weather conditions were proposed finally. The purpose of introducing the UAV platform was to cooperate with the satellite to improve the monitoring response ability of OACs and non-OACs in small and micro-sized oligotrophic and mesotrophic water bodies.

关键词： Inland water quality monitoring Multi-source remote sensing clustering algorithm Band combination model Machine learning

来源：评论

学校读者我要写书评

暂无评论

Possibilistic C-means with novel image representation for image segmentation

引用

ARTIFICIAL INTELLIGENCE REVIEW 2024年第2期58卷 1-21页

作者： Cui, Hanshuai Wang, Hongjian Zeng, Wenyi Liu, Yuqing Zhao, Bo Beijing Normal Univ Sch Artificial Intelligence Beijing 100875 Peoples R China State Grid Henan Prov Elect Power Co Management Training Ctr Zhengzhou 450015 Peoples R China Beijing Normal Univ Sch Syst Sci Beijing 100875 Peoples R China

Image segmentation is the process of automatically dividing an image into several parts and extracting the relevant data and information. Compared to the traditional Fuzzy C-Means algorithm, the Possibilistic C-Means (PCM) algorithm has advantages in reducing the influence of noise on cluster center estimation. However, the PCM algorithm still shows poor clustering performance under high-intensity noise, which may lead to overlapping cluster centers. Considering the impact of neighborhood information of image pixels on the image segmentation results, this paper proposes a Vector-Based Possibilistic C-Means (VBPCM) algorithm. The algorithm incorporates neighborhood information and uses a vector representation method to describe image pixels. Additionally, an adjustable distance based on an exponential function is proposed to describe the similarity between vectors. The proposed VBPCM algorithm outperforms the conventional PCM, obtaining uplifiting gains of 4%, 2%, and 9% in Pixel Accuracy, Mean Pixel Accuracy, and Mean Intersection over Union, respectively. The experimental outputs illustrate that VBPCM algorithm can achieve more satisfactory cluster effect with high-intensity noise, further perform better in image segmentation task.

关键词： Possibilistic C-means algorithm clustering algorithm Image denoising Image representation method Image segmentation

来源：评论

学校读者我要写书评

暂无评论

Environment Fusion Routing Protocol for Wireless Sensor Networks

引用

IEEE SENSORS JOURNAL 2024年第8期24卷 13418-13430页

作者： Yang, Fengbo Sun, Qian Zhao, Zhiyao Wang, Xiaoyi Xu, Jiping Zheng, Ying Zhang, Huiyan Wang, Li Beijing Technol & Business Univ Sch Artificial Intelligence Beijing 100048 Peoples R China Beijing Lab Intelligent Environm Protect Beijing 100048 Peoples R China Beijing Inst Fash Technol Beijing 100029 Peoples R China

Wireless sensor network (WSN) is usually deployed in unattended monitoring environments, with limited energy and large-scale characteristics. Therefore, the performance of WSN is susceptible to the influence of external environment and its own energy. The existing routing protocols try to optimize energy efficiency and routing reliability from the perspective of the network itself. However, they do not take into account the environmental impact from the outside, which makes them unable to properly respond to harsh environment. Therefore, a new environment fusion routing protocol (EFRP) is proposed. Cluster heads (CHs) are selected in a probabilistic manner according to the ratio of the energy of each node and environmental factors. Nodes with excellent environment have a higher probability of being selected as CHs. Besides, the EFRP optimal path is determined according to the comprehensive potential field superimposed by the distance field, residual energy field, and environmental field. The best path can be weighed among delay, energy saving, and routing invulnerability. Therefore, the balance of network energy is realized, and the invulnerability of the network in harsh environment is improved. The experimental results show that EFRP can significantly improve the data transmission and invulnerability of the network in harsh environments.

关键词： clustering algorithm environment aware invulnerability potential field wireless sensor network (WSN)

来源：评论

学校读者我要写书评

暂无评论

Spatial logic of scientific research: an exploratory and quantitative analysis of 42 scientific research buildings sited in America and Europe

引用

ARCHITECTURAL ENGINEERING AND DESIGN MANAGEMENT 2024年第4期20卷 850-867页

作者： Zhang, Xinwei Cui, Tong Lan, Jun Shijiazhuang Tiedao Univ Sch Architecture & Art Shijiazhuang Peoples R China Univ Chinese Acad Sci Ctr Architecture Res & Design Beijing Peoples R China 17 Beierhuan East Rd Shijiazhuang Hebei Peoples R China

Since the nineteenth-century devolution of natural philosophy into disciplines, scientific activity has expanded from single-disciplined research to comprehensive research. The flow of information between different research teams makes scientific research and its places increasingly social, which is no longer limited to the closed laboratory workbench. Reviewing the literature, we found that the development of scientific research buildings experienced a transition from efficiency to openness. And the new organization of research communities and its demand for space put forward higher requirements for scientific research buildings. To deeply explore the design trends of scientific research buildings at this stage, the sample database is established using space syntax and K-means clustering algorithm software to quantitatively analyze spatial layouts of 42 scientific research buildings worldwide from the perspective of sociology and innovation, clustered into five categories. Each category shows a different tendency in terms of accessibility, visibility, and the proportion of public space, and the different spatial configurations and openness somehow reflect the attitude of the research organization or group towards the use of space.

关键词： Scientific research buildings space syntax spatial configuration clustering algorithm

来源：评论

学校读者我要写书评

暂无评论

Field Measurement and Channel Modeling Around Wailingding Island for Maritime Wireless Communication

引用

IEEE ANTENNAS AND WIRELESS PROPAGATION LETTERS 2024年第6期23卷 1934-1938页

作者： Zhang, Hanzhong Zhou, Ting Xu, Tianheng Cheng, Min Hu, Honglin Shanghai Univ Sch Microelect Shanghai 200444 Peoples R China Chinese Acad Sci Shanghai Adv Res Inst Shanghai 201210 Peoples R China CETC Potevio Sci & Technol Co Ltd Guangzhou 510310 Peoples R China

Due to the influence of sparse scattering, wave movement and variable climate, maritime wireless channel modeling is particularly difficult. In this letter, a maritime channel measurement scheme is designed and implemented. Aclustering algorithmbased channel modeling framework is further proposed to process the measurement data and build a maritime propagation model. Specifically, the clustering algorithm is introduced in the framework to adaptively discover data characteristics in complex environments, which avoids error accumulation. Numerical results illustrate that the proposed model outperforms existing empirical path loss models in describing the signal propagation at sea, which provides good guidance for designing and deploying of maritime communication systems.

关键词： Channel measurement clustering algorithm maritime channel modeling

来源：评论

学校读者我要写书评

暂无评论

Construction and Evaluation of Information Granules: From the Perspective of clustering

引用

IEEE TRANSACTIONS ON SYSTEMS MAN CYBERNETICS-SYSTEMS 2022年第3期52卷 2024-2037页

作者： Zhu, Xiubin Pedrycz, Witold Li, Zhiwu Xidian Univ Sch Electromech Engn Xian 710071 Peoples R China Univ Alberta Dept Elect & Comp Engn Edmonton AB T6R 2V4 Canada King Abdulaziz Univ Fac Engn Jeddah 21589 Saudi Arabia Macau Univ Sci & Technol Inst Syst Engn Taipa Macau Peoples R China

While granular computing has experienced rapid growth in the past decades and some milestones have been reached, a comprehensive study of the representation capabilities delivered by numeric prototypes and granular prototypes produced by different techniques still calls for comprehensive research and a comparative analysis. Well-constructed information granules are reflective of the nature of the numeric evidence and serve as backbones of granular classifiers and granular models. The objective of this study is to review a number of clustering paradigms aimed at the construction of information granules, discuss the development of granular prototypes, and conduct a comprehensive evaluation of quality of numeric prototypes and their corresponding augmentations coming in the form of granular prototypes. We have been witnessing many studies devoted to the construction of information granules, but a comparative analysis of the quality of information granules constructed on a basis of prototypes produced by different clustering algorithms is still lacking. In this regard, the review of the clustering algorithms supporting the formation of information granules and the comprehensive comparative study of their usefulness in classification and modeling tasks offered in this study make sense. This will promote the usage of information granules in various future works, especially classification problem and system modeling.

关键词： Prototypes clustering algorithms Systems modeling Prediction algorithms Numerical models Fuzzy sets Computational modeling clustering algorithm granular computing (GrC) information granule principle of justifiable granularity

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：