检索结果-内蒙古大学图书馆

A deep learning object detection method to improve cluster analysis of two-dimensional data

MULTIMEDIA TOOLS AND APPLICATIONS 2024年第28期83卷 71171-71187页

作者： Couturier, Raphael Gregori, Pablo Noura, Hassan Salman, Ola Sider, Abderrahmane Univ Franche Comte UFC FEMTO ST Inst Besancon France Univ Jaume I Castellon Dept Matemat Inst Matemat & Aplicac Castellon E-12071 Castellon de La Plana Spain Amer Univ Beirut Elect & Comp Engn Dept Beirut Lebanon Univ Bejaia Fac Exact Sci LIMED Lab Bejaia 06000 Algeria

clustering is an unsupervised machine learning method grouping data samples into clusters of similar objects, used as a system support tool in numerous applications such as banking customers profiling, document retrieval, image segmentation, and e-commerce recommendation engines. The effectiveness of several clustering techniques is sensible to the initialization parameters, and different solutions have been proposed in the literature to overcome this limitation. They require high computational memory consumption when dealing with big data. In this paper, we propose the application of a recent object detection Deep Learning model (YOLO-v5) for assisting the initialization of classical techniques and improving their effectiveness on two-variate datasets, leveraging the accuracy and reducing dramatically the memory and time consumption of classical clustering methods.

关键词： clustering algorithms clustering initialization methods clustering initialization metrics Deep learning object detection model

来源：评论

学校读者我要写书评

暂无评论

Multimodal Moore-Penrose Inverse-Based Recomputation Framework for Big Data Analysis

引用

IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS 2024年第5期35卷 6570-6582页

作者： Zhang, Wandong Yang, Yimin Wu, Q. M. Jonathan Wang, Tianlei Zhang, Hui Univ Western Ontario Dept Elect & Comp Engn London ON N6A 3K7 Canada Univ Windsor Dept Elect & Comp Engn Windsor ON N9B 3P4 Canada Vector Inst Artificial Intelligence Toronto ON M5G 1M1 Canada Hangzhou Dianzi Univ Artificial Intelligence Inst Hangzhou 310005 Peoples R China Hunan Univ Sch Robot Changsha 410082 Peoples R China

Most multilayer Moore-Penrose inverse (MPI)-based neural networks, such as deep random vector functional link (RVFL), are structured with two separate stages: unsupervised feature encoding and supervised pattern classification. Once the unsupervised learning is finished, the latent encoding is fixed without supervised fine-tuning. However, in complex tasks such as handling the ImageNet dataset, there are often many more clues that can be directly encoded, while unsupervised learning, by definition, cannot know exactly what is useful for a certain task. There is a need to retrain the latent space representations in the supervised pattern classification stage to learn some clues that unsupervised learning has not yet been learned. In particular, the residual error in the output layer is pulled back to each hidden layer, and the parameters of the hidden layers are recalculated with MPI for more robust representations. In this article, a recomputation-based multilayer network using Moore-Penrose inverse (RML-MP) is developed. A sparse RML-MP (SRML-MP) model to boost the performance of RML-MP is then proposed. The experimental results with varying training samples (from 3k to 1.8 million) show that the proposed models provide higher Top-1 testing accuracy than most representation learning algorithms. For reproducibility, the source codes are available at https://***/W1AE/Retraining.

关键词： Representation learning Training Encoding Nonhomogeneous media Feature extraction clustering algorithms Unsupervised learning Moore-Penrose inverse (MPI) multilayer fully connected neural network multimodal encoding representation learning

来源：评论

学校读者我要写书评

暂无评论

Graph Representation Learning via Contrasting Cluster Assignments

引用

IEEE TRANSACTIONS ON COGNITIVE AND DEVELOPMENTAL SYSTEMS 2024年第3期16卷 912-922页

作者： Zhang, Chun-Yang Yao, Hong-Yu Chen, C. L. Philip Lin, Yue-Na Fuzhou Univ Coll Comp & Data Sci Fuzhou 350002 Peoples R China South China Univ Technol Sch Comp Sci & Engn Guangzhou 510006 Guangdong Peoples R China

With the rise of contrastive learning, unsupervised graph representation learning (GRL) has shown strong competitiveness. However, existing graph contrastive models typically either focus on the local view of graphs or take simple considerations of both global and local views. This may cause these models to overemphasize the importance of individual nodes and their ego networks, or to result in poor learning of global knowledge and affect the learning of local views. Additionally, most GRL models pay attention to topological proximity, assuming that nodes that are closer in graph topology are more similar. However, in the real world, close nodes may be dissimilar, which makes the learned embeddings incorporate inappropriate messages and thus lack discrimination. To address these issues, we propose a novel unsupervised GRL model by contrasting cluster assignments, called graph representation learning model via contrasting cluster assignment (GRCCA). To comprehensively explore the global and local views, it combines multiview contrastive learning and clustering algorithms with an opposite augmentation strategy. It leverages clustering algorithms to capture fine-grained global information and explore potential relevance between nodes in different augmented perspectives while preserving high-quality global and local information through contrast between nodes and prototypes. The opposite augmentation strategy further enhances the contrast of both views, allowing the model to excavate more invariant features. Experimental results show that GRCCA has strong competitiveness compared to state-of-the-art models in different graph analysis tasks.

关键词： Prototypes clustering algorithms Representation learning Unsupervised learning Task analysis Topology Social networking (online) Contrastive learning graph data mining graph representation learning (GRL) unsupervised learning

来源：评论

学校读者我要写书评

暂无评论

Effectively clustering Single Cell RNA Sequencing Data by Sparse Representation

引用

IEEE-ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS 2022年第6期19卷 3425-3434页

作者： Li, Rui-Yi Wang, Zhiye Guan, Jihong Zhou, Shuigeng Shanghai East Hosp Translat Med Ctr Stem Cell Therapy Shanghai 200136 Peoples R China Tongji Univ Sch Med Shanghai 200120 Peoples R China Tongji Univ Dept Comp Sci & Technol Shanghai 201804 Peoples R China Fudan Univ Shanghai Key Lab Intelligent Informat Proc Shanghai 200433 Peoples R China Fudan Univ Sch Comp Sci Shanghai 200433 Peoples R China

clustering analysis has been widely used in analyzing single-cell RNA-sequencing (scRNA-seq) data to study various biological problems at cellular level. Although a number of scRNA-seq data clustering methods have been developed, most of them evaluate the similarity of pairwise cells while ignoring the global relationships among cells, which sometimes cannot effectively capture the latent structure of cells. In this paper, we propose a new clustering method SPARC for scRNA-seq data. The most important feature of SPARC is a novel similarity metric that uses the sparse representation coefficients of each cell in terms of the other cells to measure the relationships among cells. In addition, we develop an outlier detection method to help parameter selection in SPARC. We compare SPARC with nine existing scRNA-seq data clustering methods on twelve real datasets. Experimental results show that SPARC achieves the state of the art performance. By further analyzing the cell similarity data derived from sparse representations, we find that SPARC is much more effective in mining high quality clusters of scRNA-seq data than two traditional similarity metrics. In conclusion, this study provides a new way to effectively cluster scRNA-seq data and achieves more accurate clustering results than the state of art methods.

关键词： Sparse matrices Data models clustering methods Computational modeling clustering algorithms Matrix decomposition Optimization Single cell RNA-seq data unsupervised sparse representation spectral clustering outlier detection

来源：评论

学校读者我要写书评

暂无评论

Research on improvement of Fuzzy K-Modes clustering algorithm for categorical data

Research on improvement of Fuzzy K-Modes clustering algorith...

引用

2024 International Conference on Image Processing and Artificial Intelligence, ICIPAl 2024

作者： Wang, Jing Yu, Liqin Si, Pinyin Wu, Guangyue Yang, Zibo School of Information Engineering Beijing Institute of Graphic Communication Beijing China

ISBN: (纸本)9781510681514

Cluster analysis is an important method in data mining and has been widely used in various aspects of real life. The Fuzzy K-Modes algorithm is one of the most commonly used algorithms for clustering categorical data due to its simplicity and effectiveness. However, this algorithm has three main problems: sensitivity of initial cluster centers, vulnerability of distance measurement methods to noise, and failure to consider different contributions of attributes during the clustering process. In this paper, we overview some existing improvement algorithms based on the three problems in detail and conduct experimental analysis of these algorithms on real data sets. Furthermore, we summarize the challenges faced in the process of improving the Fuzzy K-Modes algorithm and give possible research directions in future. © 2024 SPIE.

关键词： clustering algorithms

来源：评论

学校读者我要写书评

暂无评论

Generalized Nash Equilibrium Seeking for Directed Nonsmooth Multicluster Games via a Distributed Lipschitz Algorithm

引用

IEEE TRANSACTIONS ON CONTROL OF NETWORK SYSTEMS 2024年第4期11卷 2033-2042页

作者： Wei, Yue Zeng, Xianlin Fang, Hao Ding, Yulong Ding, Shuxin Guangdong Lab Artificial Intelligence & Digital Ec Shenzhen 518107 Peoples R China Beijing Inst Technol Key Lab Intelligent Control & Decis Complex Syst Beijing 100081 Peoples R China Tongji Univ Dept Control Sci & Engn Shanghai 200092 Peoples R China China Acad Railway Sci Corp Ltd Signal & Commun Res Inst Beijing 100081 Peoples R China

This article investigates a generalized Nash equilibrium (GNE) seeking strategy for a class of nonsmooth multicluster games. Each cluster consists of several players. The intercluster graph is directed and weight-unbalanced. Moreover, in contrast to previous works of multicluster games, coupled nonsmooth inequality constraints, resource allocation constraints, and nonsmooth payoff functions are considered simultaneously in these multicluster games. For seeking the GNE of these games, a distributed Lipschitz algorithm with the proximal-splitting scheme is proposed. Then, convergence analysis of this designed algorithm is deduced based on the Lyapunov stability theory and the convex optimization theory. Finally, some simulation results are provided in this article, which show the efficacy of the distributed GNE seeking algorithm.

关键词： Games clustering algorithms Directed graphs Heuristic algorithms Convergence Resource management Linear matrix inequalities Distributed algorithms distributed generalized Nash equilibrium (GNE) seeking multicluster games nonsmooth functions proximal operator

来源：评论

学校读者我要写书评

暂无评论

Mining Both Commonality and Specificity From Multiple Documents for Multi-Document Summarization

引用

IEEE ACCESS 2024年 12卷 54371-54381页

作者： Ma, Bing Chinese Acad Sci Inst Comp Technol Key Lab Intelligent Informat Proc Beijing 100190 Peoples R China Univ Chinese Acad Sci Beijing 101408 Peoples R China

The multi-document summarization task requires the designed summarizer to generate a short text that covers the important information of original multiple documents and satisfies content diversity. To fulfill the dual requirements of coverage and diversity in multi-document summarization, this study introduces a novel method. Initially, a class tree is constructed through hierarchical clustering of documents. Subsequently, a sentence selection method based on class tree is proposed for generating a summary. Specifically, a top-down traversal is performed on the class tree, during which sentences are selected from each node based on their similarity to the centroid of the documents within the node and their dissimilarity to the centroid of documents not belonging to the node. Sentences selected from the root node reflect the commonality of all document, and sentences selected from the sub nodes reflect the distinct specificity of the respective subclasses. Experimental results on standard text summarization datasets DUC'2002, DUC'2003, and DUC'2004 demonstrate that the proposed method significantly outperforms the variant method that considers only commonality of all documents, achieving average improvements of up to 1.54 and 1.42 in ROUGE-1 and ROUGE-L scores, respectively. Additionally, the method demonstrates significant superiority over another variant method that considers only the specificity of subclasses, achieving average improvements of up to 2.16 and 2.01 in ROUGE-1 and ROUGE-L scores, respectively. Furthermore, extensive experiments on DUC'2004 and Multi-News datasets show that the proposed method outperforms lots of competitive supervised and unsupervised multi-document summarization methods and yields considerable results.

关键词： Vectors Task analysis Feature extraction Natural languages clustering algorithms Symmetric matrices Standards Class tree commonality and specificity hierarchical clustering of documents multi-document summarization pre-trained embedding representation

来源：评论

学校读者我要写书评

暂无评论

TSF-DBSCAN: A Novel Fuzzy Density-Based Approach for clustering Unbounded Data Streams

引用

IEEE TRANSACTIONS ON FUZZY SYSTEMS 2022年第3期30卷 623-637页

作者： Bechini, Alessio Marcelloni, Francesco Renda, Alessandro Univ Pisa Dept Informat Engn I-56123 Pisa Italy Univ Florence Dept Informat Engn I-50121 Florence Italy

In recent years, several clustering algorithms have been proposed with the aim of mining knowledge from streams of data generated at a high speed by a variety of hardware platforms and software applications. Among these algorithms, density-based approaches have proved to be particularly attractive, thanks to their capability of handling outliers and capturing clusters with arbitrary shapes. The streaming setting poses additional challenges that need to be addressed as well: data streams are potentially unbounded and affected by concept drift, i.e., a modification over time in the underlying data generation process. In this article, we propose temporal streaming fuzzy density-based spatial clustering of applications with noise (TSF-DBSCAN), a novel fuzzy clustering algorithm for streaming data. TSF-DBSCAN is an extension of the well-known DBSCAN algorithm, one of the most popular density-based clustering approaches. Fuzziness is introduced in TSF-DBSCAN to model the uncertainty about the distance threshold that defines the neighborhood of an object. As a consequence, TSF-DBSCAN identifies clusters with fuzzy overlapping borders. A fading model, which makes objects less relevant as they become more remote in time, endows TSF-DBSCAN with the capability of adapting to evolving data streams. The integration of the model in a two-stage approach ensures computational and memory efficiency: during the online stage, continuously arriving objects are organized in proper data structures that are later exploited in the offline stage to determine a fine-grained partition. An extensive experimental analysis on synthetic and real-world datasets shows that TSF-DBSCAN yields competitive performance when compared to other clustering algorithms recently proposed for streaming data.

关键词： clustering algorithms Partitioning algorithms Shape Proposals Fading channels Data models Adaptation models Data stream clustering density-based clustering density-based spatial clustering of applications with noise (DBSCAN) fuzzy clustering streaming data

来源：评论

学校读者我要写书评

暂无评论

An improved multi-omics data clustering algorithm based on graph regularization constraints 3

An improved multi-omics data clustering algorithm based on g...

引用

3rd International Conference on Biomedical and Intelligent Systems, IC-BIS 2024

作者： Zhang, Shunqin Kong, Wei College of Information Engineering Shanghai Maritime University 1550 Haigang Ave. Shanghai201306 China

ISBN: (纸本)9781510681279

Accurate clustering of cell type classification by integrating multi-omics data is necessary for subsequent data analysis. However, related algorithms face difficulties in dealing with noise and missing values present in multi-omics data. The network-based integrative clustering algorithm (NIC) can overcome the problem of data heterogeneity but does not fully utilize the spatial structure information in the data. To solve this problem, this study proposes a new algorithm-graph regularization-based NIC (GR-NIC) to improve the integration accuracy, remove redundant features, and preserve the geometric structure of the data by fusing single-cell multi-omics data, single-cell sequencing data (scRNA-seq or scATACseq) and histone modification data. To verify the effectiveness of the GR-NIC algorithm, this paper compares three stateof- the-art multi-omics clustering algorithms through five datasets. The results show that GR-NIC can select features that are more representative of the structure of the data and improve the accuracy of clustering. This new approach is effectively validated on mouse brain and peripheral blood sample datasets and shows significant improvements. © 2024 SPIE.

关键词： clustering algorithms

来源：评论

学校读者我要写书评

暂无评论

clustering-FFT Based Doppler-Shift Acquisition for Space Communications

引用

IEEE TRANSACTIONS ON COMMUNICATIONS 2022年第2期70卷 1168-1182页

作者： Zhang, Zhaowei Liu, Hui Yin, Jun Shi, Xiaoye Nanjing Univ Posts & Telecommun Jiangsu Key Lab Broadband Wireless Commun Nanjing 210003 Peoples R China Nanjing Univ Posts & Telecommun Sch Internet Things Nanjing 210003 Peoples R China Jinling Inst Technol Sch Software Engn Nanjing 211169 Peoples R China

For space communications, the transmitted signal generally faces a long-distance transmission and a high-accelerating movement. The long-distance transmission causes a very low signal-to-noise ratio (SNR) and the high-accelerating movement causes a dynamic Doppler-shift. Under a very large acceleration, the long-time accumulation which aims to solve the very low SNR, brings about the serious energy dispersion problem, thus posing a great challenge for Doppler-shift acquisition. Introducing the clustering idea in machine learning, we propose one clustering-fast-Fourier-transform (CFFT) based Doppler-shift acquisition scheme with an affordable computational complexity, to address the energy dispersion problem for space communications. The proposed CFFT scheme includes two algorithms called the Generic CFFT algorithm and the High-order CFFT algorithm, respectively. First, the Generic CFFT algorithm clusters multiple density-reachable signal elements into one signal cluster, thus accumulating the dispersed energy again. Second, the High-order CFFT algorithm, which conducts several rounds of clustering, clusters more density-reachable signal elements into one larger signal cluster. Simulations results show that our proposed CFFT scheme achieves a higher acquisition probability than the existing FFT based schemes and consumes a lower computational complexity than the FRFT scheme.

关键词： Signal to noise ratio Dispersion Time-frequency analysis Space communications Receivers Computational complexity clustering algorithms Doppler-shift acquisition energy dispersion clustering machine learning FFT

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：