A Novel Multilayer Data Clustering Framework based on Feature Selection and Modified K-Means Algorithm

Duan, Ganglong; Hu, Wang; Zhang, Zhiguang

doi:10.14257/ijsip.2016.9.4.08

Cited by 4 publications

(4 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…FCM is a representative method of fuzzy clustering technique which is based on K-means to partition the dataset into clusters [10,80]. FCM algorithm is considered as a "soft" clustering method (i.e., the object is assigned to a cluster with a degree of belief) [81].…”

Section: Fuzzy C-meansmentioning

confidence: 99%

Clustering Algorithms: Taxonomy, Comparison, and Empirical Analysis in 2D Datasets

Mostafa¹

2020

Journal on Artificial Intelligence

View full text Add to dashboard Cite

Because of the abundance of clustering methods, comparing between methods and determining which method is proper for a given dataset is crucial. Especially, the availability of huge experimental datasets and transactional and the emerging requirements for data mining and the like needs badly for clustering algorithms that can be applied in various domains. This paper presents essential notions of clustering and offers an overview of the significant features of the most common representative clustering algorithms of clustering categories presented in a comparative way. More specifically the study is based on the numerical type of the data that the algorithm supports, the shape of the clusters, and complexity. The experiments were done using nine clustering algorithms representing the common clustering categories on eight 2D clustered datasets differ in the clusters' shapes and density of the data points. Furthermore, the comparison was done from the point of view seven performance measures.

show abstract

Section: Fuzzy C-meansmentioning

confidence: 99%

Clustering Algorithms: Taxonomy, Comparison, and Empirical Analysis in 2D Datasets

Mostafa¹

2020

Journal on Artificial Intelligence

View full text Add to dashboard Cite

show abstract

“…The majority of batch-clustering frameworks for multilayer networks has been built on popular learning tools such as subspace learning [37,38], fuzzy clustering [39], the wavelet transform [40], tensor decompositions [41,42], multilayer modularity maximization [16], and graph signal processing [43]. Batch approaches for multilayer networks include also [9,44,45], with [45] being able to address both state clustering and community detection, but not subnetworksequence clustering since inter-layer information cannot be accommodated.…”

Section: B Prior Artmentioning

confidence: 99%

Fast Sequential Clustering in Riemannian Manifolds for Dynamic and Time-Series-Annotated Multilayer Networks

Ye¹,

Slavakis²,

Nakuci³

et al. 2020

Preprint

View full text Add to dashboard Cite

This work exploits Riemannian manifolds to build a sequential-clustering framework able to address a wide variety of clustering tasks in dynamic multilayer (brain) networks via the information extracted from their nodal time-series. The discussion follows a bottom-up path, starting from feature extraction from time-series and reaching up to Riemannian manifolds (feature spaces) to address clustering tasks such as state clustering, community detection (a.k.a. network-topology identification), and subnetwork-sequence tracking. Kernel autoregressive-moving-average modeling and kernel (partial) correlations serve as case studies of generating features in the Riemannian manifolds of Grassmann and positive-(semi)definite matrices, respectively. Feature point-clouds form clusters which are viewed as submanifolds according to Riemannian multi-manifold modeling. A novel sequential-clustering scheme of Riemannian features is also established: feature points are first sampled in a non-random way to reveal the underlying geometric information, and, then, a fast sequential-clustering scheme is brought forth that takes advantage of Riemannian distances and the angular information on tangent spaces. By virtue of the landmark points and the sequential processing of the Riemannian features, the computational complexity of the framework is rendered free from the length of the available time-series data. The effectiveness and computational efficiency of the proposed framework is validated by extensive numerical tests against several state-of-the-art manifold-learning and brain-network-clustering schemes on synthetic as well as real functional-magnetic-resonance-imaging (fMRI) and electro-encephalogram<br> (EEG) data.

show abstract

“…One of the most important phases in data mining is cluster analysis. In order to cluster, some multi-objective algorithms can be used which automatically partition the data [1]. It should be noted that data mining (DM) consists of a set of computational techniques which are applied to discover knowledge, hidden patterns and rules obtained from data in various sciences [2].…”

Section: Introductionmentioning

confidence: 99%

“…Weights and the cluster assignments were treated together, through an iterative process. Duan et al [1] have proposed a different multilayer data clustering framework based on feature selection and modified the K-means algorithm. They have attempted to reduce the dimension of the data set, by selecting an envoy feature subset as a result of this the clustering process.…”

Section: Introductionmentioning

confidence: 99%

A novel selective clustering framework for appropriate labeling of the clusters based on K-means algorithm

2019

View full text Add to dashboard Cite

Data mining is a powerful new technology to extract hidden information from data warehouses. Data mining analyzes data from different perspectives and finds useful patterns and knowledge from large volumes of raw data. Clustering is one of the main methods of data mining. K-means algorithm is one of the most common clustering algorithms due to its efficiency and ease of use. One of the challenges of clustering is to identify the appropriate label for each cluster. The selection of a label is done in such a way as to provide a proper description of the cluster records. In some cases, choosing the appropriate label is not easy due to the results and structure of each cluster. The aim of this study is to present an algorithm based on the K-means clustering in order to facilitate the allocation of labels to each cluster.

show abstract

A Novel Multilayer Data Clustering Framework based on Feature Selection and Modified K-Means Algorithm

Abstract: Abstract

Cited by 4 publications

References 13 publications

Clustering Algorithms: Taxonomy, Comparison, and Empirical Analysis in 2D Datasets

Clustering Algorithms: Taxonomy, Comparison, and Empirical Analysis in 2D Datasets

Fast Sequential Clustering in Riemannian Manifolds for Dynamic and Time-Series-Annotated Multilayer Networks

A novel selective clustering framework for appropriate labeling of the clusters based on K-means algorithm

Contact Info

Product

Resources

About