检索结果-内蒙古大学图书馆

Implementation of FAST clustering-based Feature Subset Selection Algorithm for High-Dimensional Data 1

International Conference on Information and Communication Technology for Sustainable Development (ICT4SD)

作者： Shilu, Smit Sheth, Kushal Mehul, Ekata Charusat Univ EiTRA EInfochips Training Changa India Res Acad Ahmadabad Gujarat India

ISBN: (数字)9789811001352

ISBN: (纸本)9789811001352;9789811001338

In feature selection, we are concerned with finding out those features that produces result similar to those of the original entire set of features. We concern ourselves with efficiency and effectiveness while evaluating Feature selection algorithms. Efficiency deals with the time required to find a subset of features and effectiveness, with the quality of subset of features. On these criteria, a FAST clustering-based feature selection algorithm (FAST) has been proposed and experimentally evaluated and implemented in this paper. The dimensionality reduction of data is the most important feature of FAST. First, we use graph-theoretic clustering method to divide features into clusters. Next, we form a subset of features by selecting the feature which is most representative and strongly related to the target classes. Due to features in different clusters being relatively independent;the clustering-based strategy of FAST has a high probability of providing us with a subset of features which are both useful and independent. Efficiency of FAST is ensured by using the concept of minimum spanning tree (MST) along with kruskal's algorithm.

关键词： Feature subset selection Feature clustering Filter method Kruskal's algorithm graph-based clustering

来源：评论

学校读者我要写书评

暂无评论

Developments on Solutions of the Normalized-Cut-clustering Problem Without Eigenvectors 15th

Developments on Solutions of the Normalized-Cut-Clustering P...

引用

15th International Symposium on Neural Networks (ISNN)

作者： Leonardo Lorente-Leyva, Leandro David Herrera-Granda, Israel Rosero-Montalvo, Paul D. Ponce-Guevara, Karina L. Eduardo Castro-Ospina, Andres Becerra, Miguel A. Hernan Peluffo-Ordonez, Diego Luis Rodriguez-Sotelo, Jose Univ Tecn Norte Ibarra Ecuador Inst Tecnol Super 17 Julio Ibarra Ecuador Univ Fed Pernambuco Recife PE Brazil Inst Tecnol Metropolitano Medellin Colombia Yachay Tech Urcuqui Ecuador Corp Univ Autonoma Narino Pasto Colombia Univ Autonoma Manizales Manizales Colombia

ISBN: (纸本)9783319925370;9783319925363

Normalized-cut clustering (NCC) is a benchmark graph-based approach for unsupervised data analysis. Since its traditional formulation is a quadratic form subject to orthogonality conditions, it is often solved within an eigenvector-based framework. Nonetheless, in some cases the calculation of eigenvectors is prohibitive or unfeasible due to the involved computational cost - for instance, when dealing with high dimensional data. In this work, we present an overview of recent developments on approaches to solve the NCC problem with no requiring the calculation of eigenvectors. Particularly, heuristic-search and quadratic-formulation-based approaches are studied. Such approaches are elegantly deduced and explained, as well as simple ways to implement them are provided.

关键词： Eigenvectors graph-based clustering Normalized cut clustering Quadratic forms

来源：评论

学校读者我要写书评

暂无评论

Multi-view clustering: A Brief Review 3

Multi-view Clustering: A Brief Review

引用

3rd International Conference on Robotics and Automation Engineering (ICRAE)

作者： Tang, Zhongyang Guangdong Univ Technol Sch Comp Sci Guangzhou Guangdong Peoples R China

ISBN: (纸本)9781538695944

With the rapid development of information technology, data exhibit multi-view characteristics, and particularly single-view data cannot comprehensively describe the information of all examples. It is very significant to sufficiently make good use of the information from different views. A good multi-view learning strategy may lead to performance improvements. Multi-view clustering is one of the important branches in multi-view learning. The key problem is how to use information effectively from multiple different views, so as to discover the underlying structure of data more accurately. In general, it uses the complementary information available to improve clustering performance in multiple views. Meanwhile, it needs to ensure that the clustering results are consistent. In this paper, we review a number of representative multi-view clustering approaches in the different fields, which can be classified into four groups: cooperative style approaches, graph-based style approaches, multiple kernel learning-based approaches, and subspace learning-based approaches. Therefore, we try to summarize and analyze the development of multi-view clustering. Finally, we point out some specific challenges that are expected to improve further research in this rapidly developing field.

关键词： multi-view clustering cooperative clustering graph-based clustering multiple kernel-based clustering subspace clustering

来源：评论

学校读者我要写书评

暂无评论

graph Filtering and Automatic Parameter Selection for Efficient Spectral clustering

Graph Filtering and Automatic Parameter Selection for Effici...

引用

作者： Alshammari, Mashaan University of Sydney

学位级别：博士

Artificial intelligence agents use pattern recognition systems to carry out tasks assigned to them. Face recognition and voice analysis are examples of these tasks. Recognising a pattern can be done in a supervised or an unsupervised manner and known patterns are used to train a supervised recognition system. However, some questions can arise during the training process such as: how many known patterns should we use for training or which patterns should we label as known patterns? These questions make unsupervised pattern recognition more appealing. Numerous methods fall under unsupervised pattern recognition, which is usually referred to as a clustering method. Perhaps the best-known clustering method is k-means, which attempts to find the mean of each cluster through iterative training. In the same way, Gaussian mixture model (GMM) tries to associate points with a set of Gaussian distributions. The problem with these methods is the convexity assumption. They assume clusters are convex and share the same mean. This does not apply to simple shapes like points forming multiple lines or nested rings. The limitations of clustering, like k-means and GMMs, led researchers to develop a new method called spectral clustering. The idea behind spectral clustering is simple: instead of looking for convex clusters, let's find clusters that are strongly connected. A similarity measure is computed for each pair of points and its value is assigned as a weight on the edge connecting them. Once graph edges are computed, the top eigenvectors of the graph Laplacian form an embedding space. In that space, points that are strongly connected form convex clusters and can be detected using k-means. Despite being an effective method to detect non-convex clusters, spectral clustering has two deficiencies that made it less attractive for the pattern recognition community. First, the graph Laplacian has to pass through eigen-decomposition to find the embedding space. This has been proved to be

关键词： spectral clustering graph-based clustering

来源：评论

学校读者我要写书评

暂无评论

Parameterizing Edge Modification Problems Above Lower Bounds

引用

THEORY OF COMPUTING SYSTEMS 2018年第3期62卷 739-770页

作者： van Bevern, Rene Froese, Vincent Komusiewicz, Christian Novosibirsk State Univ Novosibirsk Russia Russian Acad Sci Sobolev Inst Math Siberian Branch Novosibirsk Russia Tech Univ Berlin Berlin Germany Friedrich Schiller Univ Jena Jena Germany

We study the parameterized complexity of a variant of the F-free Editing problem: Given a graph G and a natural number k, is it possible to modify at most k edges in G so that the resulting graph contains no induced subgraph isomorphic to F? In our variant, the input additionally contains a vertex-disjoint packing of induced subgraphs of G, which provides a lower bound on the number of edge modifications required to transform G into an F-free graph. While earlier works used the number k as parameter or structural parameters of the input graph G, we consider instead the parameter , that is, the number of edge modifications above the lower bound . We develop a framework of generic data reduction rules to show fixed-parameter tractability with respect to a"" for K (3)-Free Editing, Feedback Arc Set in Tournaments, and Cluster Editing when the packing contains subgraphs with bounded solution size. For K (3)-Free Editing, we also prove NP-hardness in case of edge-disjoint packings of K (3)s and a"" = 0, while for K (q) -Free Editing and q ae 6, NP-hardness for a"" = 0 even holds for vertex-disjoint packings of K (q) s. In addition, we provide NP-hardness results for F-free Vertex Deletion, were the aim is to delete a minimum number of vertices to make the input graph F-free.

关键词： NP-hard problem Fixed-parameter algorithm Subgraph packing Kernelization graph-based clustering Feedback arc set Cluster editing

来源：评论

学校读者我要写书评

暂无评论

Dental diagnosis from X-Ray images: An expert system based on fuzzy computing

引用

BIOMEDICAL SIGNAL PROCESSING AND CONTROL 2018年第0期39卷 64-73页

作者： Le Hoang Son Tran Manh Tuan Fujita, Hamido Dey, Nilanjan Ashour, Amira S. Vo Truong Nhu Ngoc Le Quynh Anh Dinh-Toi Chu Vietnam Natl Univ VNU Univ Sci Hanoi Vietnam Thuy Loi Univ Hanoi Vietnam Iwate Prefectural Univ Takizawa Iwate Japan Techno India Coll Technol Kolkata India Tanta Univ Fac Engn Tanta Egypt Hanoi Med Univ Sch Odontostomatol Hanoi Vietnam Duy Tan Univ Inst Res & Dev 03 Quang Trung Danang Vietnam Hanoi Natl Univ Educ Fac Biol Hanoi Vietnam

Background: Computerized medical diagnosis systems from X-Ray images are of great interest to physicians for accurate decision making of possible diseases and treatments. Subclinical disease has no recognizable clinical findings, thus it is desirable to segment the dental X-Ray image into groups and then use soft computing methods to check the possibility of whether or not any disease occurs therein. Methods: The current work proposed a novel framework called Dental Diagnosis System (DDS) for dental diagnosis based on the hybrid approach of segmentation, classification and decision making. It utilized the best dental image segmentation method based on semi-supervised fuzzy clustering for the segmentation task A new graph-based clustering algorithm called APC+ for the classification task was proposed. A new decision making procedure was designed to determine the final disease from a group of diseases found from the segments. Results: The proposed DDS was modeled under the real dental case of Hanoi Medical University, Vietnam including 87 dental images of five popular diseases, namely: root fracture, incluse teeth, decay, missing teeth, and resorption of periodontal bone. The DDS accuracy is 92.74% which is superior to the other methods namely fuzzy inference system (89.67%), fuzzy k-nearest neighbor (80.05%), prim spanning tree (58.46%), kruskal spanning tree (58.46%), and affinity propagation clustering (90.01%). Conclusion: Empirical-results established that superior performance of the DDS to other related methods the findings of the achieved results can assist dental clinicians in their professional work. (C) 2017 Elsevier Ltd. All rights reserved.

关键词： Dental X-Ray image Decision making graph-based clustering Medical diagnosis Semi-Supervised fuzzy clustering

来源：评论

学校读者我要写书评

暂无评论

Hierarchical inference of unicast network topologies based on end-to-end measurements

引用

IEEE TRANSACTIONS ON SIGNAL PROCESSING 2007年第5期55卷 1708-1718页

作者： Shih, Meng-Fu Hero, Alfred O., III Univ Michigan Dept Elect Engn & Comp Sci Ann Arbor MI 48109 USA

In this paper, we address the problem of topology discovery in unicast logical tree networks using end-to-end measurements. Without any cooperation from the internal routers, topology estimation can be formulated as hierarchical clustering of the leaf nodes based on pairwise correlations as similarity metrics. Unlike previous work that first assumes the network topology is a binary tree and then tries to generalize to a nonbinary tree, we provide a framework that directly deals with general logical tree topologies. A hierarchical algorithm to estimate the topology is developed in a recursive manner by finding the best partitions of the leaf nodes level by level. Our simulations show that the algorithm is more robust than binary-tree based methods.

关键词： graph-based clustering mixture models network tomography topology estimation

来源：评论

学校读者我要写书评

暂无评论

A framework for crime data analysis using relationship among named entities

引用

NEURAL COMPUTING & APPLICATIONS 2020年第12期32卷 7671-7689页

作者： Das, Priyanka Das, Asit Kumar Nayak, Janmenjoy Pelusi, Danilo Indian Inst Engn Sci & Technol Dept Comp Sci & Technol Sibpur 711103 Howrah India Sri Sivani Coll Engn Dept Comp Sci & Engn Chilakapalem Andhra Pradesh India Univ Teramo Dept Commun Sci Teramo Italy

Many crime reports are available online in various blogs and Newswire. Though manual annotation of these massive reports is quite tedious for crime data analysis, it gives an overall crime scenario of all over the world. This motivates us to propose a framework for crime data analysis based on the online reports. Initially, the method extracts the crime reports and identifies named entities. The intermediate sequence of context words between every consecutive pair of named entities is termed as a crime vector that provides relationships between the entities. The feature vectors for each entity pair are generated from these crime vectors using the Word2Vec model. The paper considers three different types of named entity pairs to facilitate the major crime data analysis task, and for each type, similarity between every pair of entities is measured using respective feature vectors. For each type of named entity pair, a separate weighted graph is generated with entity pairs as vertices and similarity score between them as the weight of the corresponding edge. Then, Infomap, a graph-based clustering algorithm, is applied to obtain optimal set of clusters of entity pairs and a representative entity pair of each cluster. Each cluster is labelled by the relationship, represented by the crime vector, of its representative entity pair. In reality, all the entity pairs in a cluster may not reflect contextual similarity with their representative entity pair. So the clusters are further partitioned into subclusters based on WordNet-based path similarity measure which makes the entity pairs in each subcluster more contextually similar compared to their original cluster. These subclusters provide us various statistical crime information over the time period. The method is experimented only using the crime reports related to crime against women in India. The experimental results demonstrate the effectiveness and superiority of the method compared to others for analysing the crime d

关键词： Crime analysis Online news Entity recognition Relation extraction Paraphrase extraction graph-based clustering

来源：评论

学校读者我要写书评

暂无评论

Exact algorithms for the minimum s-club partitioning problem

引用

ANNALS OF OPERATIONS RESEARCH 2019年第1-2期276卷 267-291页

作者： Yezerska, Oleksandra Pajouh, Foad Mahdavi Veremyev, Alexander Butenko, Sergiy Texas A&M Univ Ind & Syst Engn College Stn TX 77843 USA Univ Massachusetts Dept Management Sci & Informat Syst Boston MA 02125 USA Univ Cent Florida Dept Ind Engn & Management Syst 12800 Pegasus Dr Orlando FL 32816 USA

graph clustering (partitioning) is a helpful tool in understanding complex systems and analyzing their structure and internal properties. One approach for graph clustering is based on partitioning the graph into cliques. However, clique models are too restrictive and prone to errors given imperfect data. Thus, using clique relaxations instead may provide a more reasonable and applicable partitioning of the graph. An s-club is a distance-based relaxation of a clique and is formally defined as a subset of vertices inducing a subgraph with a diameter of at most s. In this work, we study the minimum s-club partitioning problem, which is to partition the graph into a minimum number of non-overlapping s-club clusters. Integer programming techniques and combinatorial branch-and-bound framework are employed to develop exact algorithms to solve this problem. We also study and compare the computational performance of the proposed algorithms for the special cases of s=2 and s=3 on a test-bed of randomly generated instances and real-life graphs.

关键词： s-Clubs Partitioning graph-based clustering Integer programming Combinatorial branch-and-bound

来源：评论

学校读者我要写书评

暂无评论

Genome-wide repeat dynamics reflect phylogenetic distance in closely related allotetraploid Nicotiana (Solanaceae)

引用

PLANT SYSTEMATICS AND EVOLUTION 2017年第8期303卷 1013-1020页

作者： Dodsworth, Steven Jang, Tae-Soo Struebig, Monika Chase, Mark W. Weiss-Schneeweiss, Hanna Leitch, Andrew R. Queen Mary Univ London Sch Biol & Chem Sci London E1 4NS England Royal Bot Gardens Dept Comparat Plant & Fungal Biol Richmond TW9 3DS Surrey England Univ Vienna Dept Systemat & Evolutionary Bot Rennweg 14 A-1030 Vienna Austria Univ Western Australia Sch Plant Biol Crawley WA 6009 Australia

Nicotiana sect. Repandae is a group of four allotetraploid species originating from a single allopolyploidisation event approximately 5 million years ago. Previous phylogenetic analyses support the hypothesis of N. nudicaulis as sister to the other three species. This is concordant with changes in genome size, separating those with genome downsizing (N. nudicaulis) from those with genome upsizing (N. repanda, N. nesophila, N. stocktonii). However, a recent analysis reflecting genome dynamics of different transposable element families reconstructed greater similarity between N. nudicaulis and the Revillagigedo Island taxa (N. nesophila and N. stocktonii), thereby placing N. repanda as sister to the rest of the group. This could reflect a different phylogenetic hypothesis or the unique evolutionary history of these particular elements. Here we re-examine relationships in this group and investigate genome-wide patterns in repetitive DNA, utilising high-throughput sequencing and a genome skimming approach. Repetitive DNA clusters provide support for N. nudicaulis as sister to the rest of the section, with N. repanda sister to the two Revillagigedo Island species. Clade-specific patterns in the occurrence and abundance of particular repeats confirm the original (N. nudicaulis (N. repanda (N. nesophila + N. stocktonii))) hypothesis. Furthermore, overall repeat dynamics in the island species N. nesophila and N. stocktonii confirm their similarity to N. repanda and the distinctive patterns between these three species and N. nudicaulis. Together these results suggest that broad-scale repeat dynamics do in fact reflect evolutionary history and could be predicted based on phylogenetic distance.

关键词： Chromoviruses graph-based clustering High-throughput sequencing Phylogenetics Repetitive DNA Ty-3 Gypsy

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：