检索结果-内蒙古大学图书馆

CLSM-FL: clustering-Based Semantic Federated Learning in Non-IID IoT Environment

IEEE INTERNET OF THINGS JOURNAL 2024年第18期11卷 29838-29851页

作者： Lee, Hunmin Seo, Daehee Univ Minnesota Dept Comp Sci & Engn Minneapolis MN 55455 USA Sangmyung Univ Dept Artificial Intelligence & Data Engn Seoul 03016 South Korea

Federated learning (FL) stands as a robust framework facilitating collaborative learning in distributed Internet of Things (IoT) settings. Nevertheless, the proliferation of nonidentically and independently distributed (non-IID) data sets presents formidable challenges, revealing distinct attributes among local participants within the IoT network. In response, clustering has been incorporated into FL methodologies, bolstering adaptability to heterogeneous data set characteristics and ameliorating the effects of non-IID conditions. Despite its effectiveness, the interpretive aspects and ramifications of clustering within FL have received insufficient scrutiny, constraining its broader utility in artificial intelligence-infused IoT ecosystems. This study exposes a notable drawback in traditional clustering methods within FL, labeled as "Cluster Collapse Disorder," wherein cluster boundaries blur, leaving models susceptible to non-IID data and data drift. Inspired by this finding, we propose clustering-based semantic FL (CLSM-FL). Our framework employs a soft clustering technique based on the inherent attributes of local models, assigning them to semantic clusters for interpretable evaluation of their capabilities. Beyond its interpretive benefits, CLSM-FL ensures resilient performance in non-IID scenarios and against data-perturbed adversarial attacks. We substantiate its efficacy through extensive experiments covering various scenarios and data sets. This contribution not only underscores the crucial role of clustering in FL but also advances comprehension and practical application of FL methodologies in non-IID IoT environments.

关键词： clustering algorithm distributed Internet of Things (IoT) environment federated learning (FL) interpretative machine learning architecture nonidentically and independently distributed (non-IID) data set optimization in distributed learning

来源：评论

学校读者我要写书评

暂无评论

A Multitask Network for Multiview Stereo Reconstruction: When Semantic Consistency-Based clustering Meets Depth Estimation Optimization

引用

IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING 2024年 62卷 1页

作者： Huang, Xin Zhang, Shulei Li, Jiayi Wang, Leiguang Wuhan Univ Sch Remote Sensing & Informat Engn Wuhan 430079 Peoples R China Hubei Luojia Lab Wuhan 430079 Peoples R China Southwest Forestry Univ Inst Big Data & Artificial Intelligence Kunming 650024 Peoples R China Southwest Forestry Univ Key Lab State Forestry & Grassland Adm Forestry & Kunming 650024 Peoples R China

We propose a novel network for multiview stereo (MVS) reconstruction in the field of remote sensing, which considers clustering-based semantic consistency into depth estimation optimization, referred to as CSC-MVS. In this approach, high-level semantic information acquired from multiple views is used to construct semantic consistency and assist in guiding the optimization of the MVS network. Specifically, the nonnegative matrix factorization (NMF) branch and the deep spectral decomposition (DSD) branch are designed to generate local and global semantic guidance, respectively. We then propose an uncertainty multitask optimization method to adaptively combine matching and semantic metrics. The performance of CSC-MVS is evaluated on representative benchmarks, including the WHU TLC dataset and LuoJia-MVS dataset, demonstrating its effectiveness and generality across diverse remote sensing scenarios. Comprehensive experimental results show that our CSC-MVS significantly improves the performance of various MVS baseline networks and achieves notable accuracy in depth reconstruction. We also conduct ablation studies to validate the rationality of each component and sensitivity analysis to confirm the robustness and adaptability of our proposed method. The code is available at https://***/zsl-whu/csc-mvs.

关键词： clustering algorithm deep learning depth map-based multiview stereo (MVS) reconstruction remote sensing imagery

来源：评论

学校读者我要写书评

暂无评论

Developing a forecasting model for time series based on clustering and deep learning algorithms

引用

APPLIED SOFT COMPUTING 2025年 174卷

作者： Luan, Nguyen-Huynh Tai, Vo-Van Ho Chi Minh City Univ Technol HCMUT Fac Appl Sci Dept Appl Math 268 Ly Thuong Kiet StDist 10 Ho Chi Minh City Vietnam Vietnam Natl Univ Ho Chi Minh City Linh Trung Ward Ho Chi Minh City Vietnam Van Lang Univ Fac Mech Elect & Comp Engn Sch Technol Ho Chi Minh City Vietnam Can Tho Univ Coll Nat Sci Can Tho Vietnam

This study proposes a new forecasting model for time series based on the improvement and combination of the cluster analysis (CA) algorithm and deep learning with Convolutional Neural Network (CNN) and Bi-Long Short Term Memory (BiLSTM) model. The proposed model is considered pioneering in this research direction with significant contributions to three main phases. For the first phase, the original series is converted into the percentage change series and is divided into clusters of an appropriate number using the CA algorithm. The next phase involves extracting the features of the new series based on the CNN with suitable parameters and input data enhancement from the results of the first phase. In the final phase, the BiLSTM model is applied to the series established from the second phase, and the forecasting principle for the future is established. The proposed model is detailed in the implementation steps, proving convergence, illustrated by numerical examples, and can be applied to real series using a Matlab procedure. The effectiveness of the proposed model is quite impressive as it surpasses many strong forecasting models on reputable benchmark datasets , including the M3-Competition dataset with 3,003 series, and M4-Competition dataset with 100,000 series.

关键词： clustering algorithm Convolutional neural network Forecasting model Long short term memory Time series

来源：评论

学校读者我要写书评

暂无评论

Scheduling for batch processes based on clustering approximated timed reachability graphs

引用

ISA TRANSACTIONS 2024年 152卷 290-298页

作者： Zhou, Jiazhong Lefebvre, Dimitri Li, Zhiwu Huaqiao Univ Coll Informat Sci & Engn Xiamen 361021 Peoples R China Xidian Univ Sch Electromech Engn Xian 710071 Peoples R China Univ La Havre Normandie Normandie Univ GREAH UR 3220 F-76600 Le Havre France Macau Univ Sci & Technol Inst Syst Engn Macau Peoples R China

To solve some scheduling problems of batch processes based on timed Petri net models, timed extended reachability graphs (TERGs) and approximated TERGs can be used. Such graphs abstract temporal specifications and represent parts of timed languages. By exploring the feasible trajectories in a TERG, optimal schedules can be obtained with respect to the makespans of batch processes that are modeled by timed Petri nets. Nevertheless, the rapid growth of the number of states in a TERG makes the problem intractable for large systems. In this paper, we improve the existing clustering TERG approach, and we make it suitable for large sized batch processes. We also enlarge a systematic approach to model batch processes with timed Petri nets. Finally, a comprehensive example of scheduling problem is studied for an archetypal chemical production plant in order to illustrate the efficiency of the proposed approach.

关键词： Batch process Timed Petri net Reachability graph clustering algorithm Scheduling

来源：评论

学校读者我要写书评

暂无评论

Evaluation of agreement between common clustering strategies for DNA methylation-based subtyping of breast tumours

引用

EPIGENOMICS 2025年第2期17卷 105-114页

作者： Zarean, Elaheh Li, Shuai Wong, Ee Ming Makalic, Enes Milne, Roger L. Giles, Graham G. Mclean, Catriona Southey, Melissa C. Dugue, Pierre-Antoine Monash Univ Monash Hlth Sch Clin Sci Precis Med 246 Clayton Rd Clayton Vic 3168 Australia Univ Melbourne Ctr Epidemiol & Biostat Melbourne Sch Populat & Global Hlth Parkville Vic Australia Monash Univ Fac Informat Technol Dept Data Sci & AI Clayton Vic Australia Canc Council Victoria Canc Epidemiol Div Melbourne Vic Australia Alfred Hlth Alfred Hosp Anat Pathol Melbourne Vic Australia Univ Melbourne Melbourne Med Sch Dept Clin Pathol Parkville Vic Australia

Aims: clustering algorithms have been widely applied to tumor DNA methylation datasets to define methylation-based cancer subtypes. This study aimed to evaluate the agreement between subtypes obtained from common clustering strategies, Materials & Methods: We used tumor DNA methylation data from 409 women with breast cancer from the Melbourne Collaborative Cohort Study (MCCS) and 781 breast tumors from The Cancer Genome Atlas (TCGA). Agreement was assessed using the adjusted Rand index for various combinations of number of CpGs, number of clusters and clustering algorithms (hierarchical, K-means, partitioning around medoids, and recursively partitioned mixture models). Results: Inconsistent agreement patterns were observed for between-algorithm and within-algorithm comparisons, with generally poor to moderate agreement (ARI <0.7). Results were qualitatively similar in the MCCS and TCGA, showing better agreement for moderate number of CpGs and fewer clusters (K = 2). Restricting the analysis to CpGs that were differentially-methylated between tumor and normal tissue did not result in higher agreement. Conclusion: Our study highlights that common clustering strategies involving an arbitrary choice of algorithm, number of clusters and number of methylation sites are likely to identify different DNA methylation-based breast tumor subtypes.

关键词： Breast tumor DNA methylation clustering algorithm agreement adjusted rand index

来源：评论

学校读者我要写书评

暂无评论

A task-driven scheme for forming clustering-structure-based heterogeneous FANETs

引用

VEHICULAR COMMUNICATIONS 2025年 52卷

作者： Chen, Siji Jiang, Bo Xu, Hong Pang, Tao Gao, Mingke Liu, Ziyang Chongqing Univ Posts & Telecommun Sch Comp Sci & Technol Chongqing 400065 Peoples R China East China Res Inst Comp Technol Shanghai 201800 Peoples R China Hangzhou Dianzi Univ Sch Comp Sci Hangzhou 310018 Peoples R China

Unmanned aerial vehicles (UAVs) are an emerging technology with the potential to be used in industries and various sectors of human life to provide a wide range of applications and services, significantly enhancing its applicability in different fields. When a UAV swarm performs complex tasks, flying Ad-hoc networks (FANETs) based on cluster structures have become a key research topic in the field of topology control due to their strong scalability and low routing overhead. However, current research mainly concentrates on the selection of the cluster head (CH), considering all UAVs within the CH's communication radius as cluster members (CMs), often neglecting whether the cluster can effectively accomplish the task, thereby potentially leading to mission failure. To overcome this problem, this paper innovatively proposes a task-driven clustering (TDC-MOPSO) algorithm based on improved multi-objective particle swarm optimization (MOPSO) for clustering-structure- based heterogeneous FANETs, which introduces the transfer function to improve the search range of particles and the mutation mechanism to avoid falling into local optima, and a more reasonable fitness function is designed to select CHs. The simulation results indicate that the proposed TDC-MOPSO algorithm dramatically improves the task completion rate by up to about 41.32% and extends the node lifetime by up to about 50.12% compared to traditional clustering algorithms. Meanwhile, the TDC-MOPSO algorithm improves the task completion rate by up to about 11.02% compared to other mopso-based algorithms. Furthermore, the TDC-MOPSO algorithm obtains more clustering solutions with higher average energy, less waste of resources, less CH handover rate, and less routing overhead in simulation. The proposed algorithm is also verified in a real-life scenario, which also effectively supports the completion of the task. All of which demonstrates that the TDC-MOPSO algorithm enhances the efficiency of task execution while

关键词： Unmanned aerial vehicles (UAVs) clustering algorithm Multi-objective particle swarm optimization (MOPSO) Flying ad-hoc networks (FANETs)

来源：评论

学校读者我要写书评

暂无评论

Plot-scale population estimation modeling based on residential plot form clustering and locational attractiveness analysis

引用

COMPUTERS ENVIRONMENT AND URBAN SYSTEMS 2025年 118卷

作者： Peng, Youmei Liu, Quan Nanjing Univ Sch Architecture & Urban Planning Nanjing 210093 Peoples R China

In many regions, urbanization has advanced to a stage that requires urban renewal, making precise population data essential for effective regional renewal and sustainable development. Therefore, this paper aims to disaggregate Jiedao-level (an administrative unit under the district) census population data to the Plot level. From an urban morphology perspective, the Gaussian Mixture Model (GMM) clustering algorithm was applied to classify the form of residential plots, assigning a type parameter for each type: the per capita housing area, to describe population density differences among the types. We then used Pearson correlation analysis to assess the relationship between POI density and population density at various bandwidths, identifying the optimal bandwidth for different POI types and calculating the overall POI density for each plot to evaluate its locational attractiveness. A regression model was established using per capita housing area, POI density, and total building area to derive population weight layers for estimating population at the plot level. The results of accuracy assessment show that using the morphological type parameter can effectively improve the estimation accuracy at plot scale, especially in areas with diverse land-use patterns and lower population density. However, our optimized locational attractiveness calculation method shows only a slight improvement to the method using a fixed bandwidth. This study develops a more accurate population estimation method of plot-level based on morphological classification, and highlights the population distribution characteristics of different types of residential plots, aiding urban decision-makers in developing targeted strategies for housing optimization and community resource allocation.

关键词： Population estimation Residential plot Points-of-interest Morphological type clustering algorithm Urban renewal

来源：评论

学校读者我要写书评

暂无评论

clustering performance analysis using a new correlation-based cluster validity index

引用

PATTERN RECOGNITION 2024年 145卷

作者： Wiroonsri, Nathakhun King Mongkuts Univ Technol Thonburi Fac Sci Dept Math Math & Stat Applicat Res Grp MaSA Bangkok Thailand King Mongkuts Univ Technol Thonburi 126 Pracha Uthit Rd Bangkok 10140 Thailand

There are various cluster validity indices used for evaluating clustering results. One of the main objectives of using these indices is to seek the optimal unknown number of clusters. Some indices work well for clusters with different densities, sizes, and shapes. Yet, one shared weakness of those validity indices is that they often provide only one optimal number of clusters. That number is unknown in real-world problems, and there might be more than one possible option. We develop a new cluster validity index based on a correlation between an actual distance between a pair of data points and a centroid distance of clusters that the two points occupy. Our proposed index constantly yields several local peaks and overcomes the previously stated weakness. Several experiments in different scenarios, including UCI real-world data sets, have been conducted to compare the proposed validity index with several well-known ones.

关键词： clustering algorithm Cluster validity measure Data partitions Correlation Pattern recognition Marketing

来源：评论

学校读者我要写书评

暂无评论

Extraction of Representative Scenarios for Photovoltaic Power With Shared Weight Graph clustering

引用

IEEE TRANSACTIONS ON SMART GRID 2024年第6期15卷 6158-6170页

作者： Fu, Xueqian Lu, Na Sun, Hongbin Zhang, Youmin China Agr Univ Coll Informat & Elect Engn Beijing 100083 Peoples R China China Agr Univ Natl Innovat Ctr Digital Fishery Beijing 100083 Peoples R China Tsinghua Univ Dept Elect Engn Beijing 100084 Peoples R China Concordia Univ Dept Mech Ind & Aerosp Engn Montreal PQ H3G 1M8 Canada

With the growing integration of photovoltaic (PV) generation, the operational conditions of power systems become more complex and variable. These intricate scenarios place significant pressure on the optimization calculations for power systems, necessitating the extraction of representative scenarios for PV power generation to enhance optimization efficiency. To address this issue, we have proposed a novel clustering model that extracts representative PV output scenarios through the fusion of adaptive feature weights and adjacent density weights. We propose an alternating optimization solution algorithm based on the Lagrange multiplier method and eigenvalue decomposition. The highlight of this work is the dual verification through theoretical proof and simulation experiment. In terms of theoretical proof, we analyze the sensitivity of clustering model parameters, demonstrate algorithm complexity, and theoretically prove the convergence of the proposed solution algorithm. Using actual PV output data from Australia, we validate the high cohesion, low coupling, noise resistance, and parameter sensitivity of the proposed clustering model, as well as the convergence of the proposed solution algorithm. The effectiveness of the proposed method in extracting representative scenarios of PV output has been confirmed through probabilistic power flow analysis using two IEEE test cases.

关键词： Power generation clustering algorithms Optimization Planning Power systems Feature extraction Uncertainty Photovoltaic representative scenario active distribution network probabilistic power flow clustering algorithm

来源：评论

学校读者我要写书评

暂无评论

A New Hierarchical clustering algorithm

A New Hierarchical Clustering Algorithm

引用

International Conference on Intelligent Informatics and Biomedical Sciences

作者： Zahra Nazari Dongshik Kang M. Reza Asharif Yulwan Sung Seiji Ogawa Graduate School of Engineering & Science University of the Ryukyus Kansei Fukushi Research Center Tohoku Fukushi University

ISBN: (纸本)9781479985630

The purpose of data clustering algorithm is to form clusters (groups) of data points such that there is high intra-cluster and low inter-cluster similarity. There are different types of clustering methods such as hierarchical, partitioning, grid and density based. Hierarchical clustering is a method of cluster analysis which seeks to build a hierarchy of clusters. A hierarchical clustering method can be thought of as a set of ordinary (flat) clustering methods organized in a tree structure. These methods construct the clusters by recursively partitioning the objects in either a top-down or bottom-up fashion. In this paper we present a new hierarchical clustering algorithm using Euclidean distance. To validate this method we have performed some experiments with low dimensional artificial datasets and high dimensional fMRI dataset. Finally the result of our method is compared to some of existing clustering methods.

关键词： clustering algorithm Hierarchical clustering algorithm Agglomerative hierarchical clustering FMRI data clustering

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：