Subspace Clustering for High-Dimensional Data Using Cluster Structure Similarity

Kavan Fatehi, Mohsen Rezvani, Mansoor Fateh and Mohammad-Reza Pajoohan
Additional contact information
Kavan Fatehi: Yazd University, Department of Computer Engineering, Yazd, Islamic Republic of Iran
Mohsen Rezvani: Shahrood University of Technology, Department of Computer Engineering, Shahrood, Islamic Republic of Iran
Mansoor Fateh: Shahrood University of Technology, Department of Computer Engineering, Shahrood, Islamic Republic of Iran
Mohammad-Reza Pajoohan: Yazd University, Department of Computer Engineering, Yazd, Islamic Republic of Iran

International Journal of Intelligent Information Technologies (IJIIT), 2018, vol. 14, issue 3, 38-55

Abstract: This article describes how recently, because of the curse of dimensionality in high dimensional data, a significant amount of research has been conducted on subspace clustering aiming at discovering clusters embedded in any possible attributes combination. The main goal of subspace clustering algorithms is to find all clusters in all subspaces. Previous studies have mostly been generating redundant subspace clusters, leading to clustering accuracy loss and also increasing the running time of the algorithms. A bottom-up density-based approach is suggested in this article, in which the cluster structure serves as a similarity measure to generate the optimal subspaces which result in raising the accuracy of the subspace clustering. Based on this idea, the algorithm discovers similar subspaces by considering similarity in their cluster structure, then combines them and the data in the new subspaces would be clustered again. Finally, the algorithm determines all the subspaces and also finds all clusters within them. Experiments on various synthetic and real datasets show that the results of the proposed approach are significantly better in quality and runtime than the state-of-the-art on clustering high-dimensional data.

Date: 2018
References: Add references at CitEc
Citations: View citations in EconPapers (2)

Downloads: (external link)
http://services.igi-global.com/resolvedoi/resolve. ... 018/IJIIT.2018070103 (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:igg:jiit00:v:14:y:2018:i:3:p:38-55

Access Statistics for this article

International Journal of Intelligent Information Technologies (IJIIT) is currently edited by Vijayan Sugumaran

More articles in International Journal of Intelligent Information Technologies (IJIIT) from IGI Global
Bibliographic data for series maintained by Journal Editor ().