Paper Info
Title
E-Means: An Evolutionary Clustering Algorithm
Abstract
In this paper we propose a new evolutionary clustering algorithm named E-means . E-means is an E volutionary extension of k-means algorithm that is composed by a revised k-means algorithm and an evolutionary approach to Gaussian mixture model, which estimates automatically the number of clusters and the optimal mean for each cluster. More specifically, the proposed E-means algorithm defines an entropy-based fitness function, and three genetic operators for merging, mutation, and deletion components. We conduct two sets of experiments using a synthetic dataset and an existing benchmark to validate the proposed E-means algorithm. The results obtained in the first experiment show that the algorithm can estimate exactly the optimal number of clusters for a set of data. In the second experiment, we compute nine major clustering validity indices and compare the corresponding results with those obtained using four established clustering techniques, and found that our E-means algorithm achieves better clustering structures.
Year
DOI
Venue
2008
10.1007/978-3-540-92137-0_59
ISICA
Keywords
Field
DocType
major clustering validity index,proposed e-means algorithm,evolutionary clustering algorithm,k-means algorithm,optimal mean,new evolutionary clustering algorithm,experiment show,revised k-means algorithm,e-means algorithm,evolutionary approach,clustering structure,gaussian mixture model,fitness function,genetic operator,k means,evolutionary computing,k means algorithm
Canopy clustering algorithm,Fuzzy clustering,CURE data clustering algorithm,Correlation clustering,Algorithm,Determining the number of clusters in a data set,Nearest-neighbor chain algorithm,Cluster analysis,Mathematics,Single-linkage clustering
Conference
Volume
ISSN
Citations 
5370
0302-9743
0
PageRank 
References 
Authors
0.34
1
3
Name
Order
Citations
PageRank
Wei Lu170330.81
Hengjian Tong241.19
Issa Traore330632.31