Integrating Fuzzy K-Means, Particle Swarm Optimization, and Imperialist Competitive Algorithm for Data Clustering

In this paper, we proposed two hybrid data clustering algorithms that are called ICAFKM and PSOFKM. ICAFKM combined the advantageous aspects of Fuzzy K-Means (FKM) and Imperialist Competitive Algorithm (ICA), and PSOFKM makes full use of the merits of both Particle Swarm Optimization (PSO) and FKM a...

Full description

Saved in:
Bibliographic Details
Published in:Arabian Journal for Science and Engineering Vol. 40; no. 12; pp. 3545 - 3554
Main Authors: Emami, Hojjat, Derakhshan, Farnaz
Format: Journal Article
Language:English
Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.12.2015
Subjects:
Engineering
Humanities and Social Sciences
multidisciplinary
Research Article - Computer Engineering and Computer Science
Science
Fuzzy K-Means algorithm
Particle Swarm Optimization
Optimization
Data clustering
Imperialist Competitive Algorithm
ISSN:1319-8025, 2191-4281
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In this paper, we proposed two hybrid data clustering algorithms that are called ICAFKM and PSOFKM. ICAFKM combined the advantageous aspects of Fuzzy K-Means (FKM) and Imperialist Competitive Algorithm (ICA), and PSOFKM makes full use of the merits of both Particle Swarm Optimization (PSO) and FKM algorithms. FKM is one of the most popular data clustering methods. However, this algorithm solves the problem of sensitivity to initial states of K-Means (KM) algorithm, but like KM, it often converges to local optima. The proposed ICAFKM and PSOFKM algorithms aim to help the FKM to escape from local optima and increase the convergence speed of the ICA and PSO algorithms in clustering process. In order to evaluate the performance of ICAFKM and PSOFKM methods, we evaluate these algorithms on five datasets and compared them with FKM, ICA, PSO, PSOKHM, and HABC algorithms. The experimental results indicate that the ICAFKM carries out better results than the other methods.
ISSN:1319-8025
2191-4281
DOI:10.1007/s13369-015-1826-3