A survey on hard subspace clustering algorithms
A. Surekha
, S. Anuradha , B. Jaya Lakshmi , K. B. Madhuri
Subspace clustering, Hard subspace clustering, Top-down approach, Bottom-up approach
Abstract---Subspace clustering is an extension to traditional clustering that seeks to find clusters in different subspaces within a dataset. Subspace clustering finds sets of objects that are homogeneous in subspaces of high-dimensional datasets, and has been successfully applied in many domains. Often in high dimensional data, many dimensions may be irrelevant and can mask real clusters. Subspace clustering algorithms localize the search process for relevant dimensions allowing them to find clusters that exist in various subspaces. Subspace clustering can be categorized into hard subspace clustering (HSC) and soft subspace clustering (SSC). HSC algorithms assume that each dimension in the data set has equal importance in the process of clustering, while SSC algorithms deal with feature weighing based on its contribution. Based on the direction of exploration of subspace clusters, HSC algorithms could be classified into two main categories: Top-down and Bottom-up. Top-down algorithms find an initial clustering in the full set of dimensions and evaluate the subspaces of each cluster, iteratively improving the results. Bottom-up approaches find dense regions in low dimensional spaces and combine them to form clusters. This paper surveys various hard subspace clustering algorithms and their efficacies, insufficiencies and recent developments. The readers would be provided with clear outline about the existing algorithms and nurture further developments and significant research in the area.
"A survey on hard subspace clustering algorithms", IJSDR - International Journal of Scientific Development and Research (www.IJSDR.org), ISSN:2455-2631, Vol.1, Issue 8, page no.321 - 326, August-2016, Available :https://ijsdr.org/papers/IJSDR1608040.pdf
Volume 1
Issue 8,
August-2016
Pages : 321 - 326
Paper Reg. ID: IJSDR_160702
Published Paper Id: IJSDR1608040
Downloads: 000347169
Research Area: Engineering
Country: Visakhapatnam, Andhra Pradesh, India
ISSN: 2455-2631 | IMPACT FACTOR: 9.15 Calculated By Google Scholar | ESTD YEAR: 2016
An International Scholarly Open Access Journal, Peer-Reviewed, Refereed Journal Impact Factor 9.15 Calculate by Google Scholar and Semantic Scholar | AI-Powered Research Tool, Multidisciplinary, Monthly, Multilanguage Journal Indexing in All Major Database & Metadata, Citation Generator
Publisher: IJSDR(IJ Publication) Janvi Wave
