Comparison Among Methods for k Estimation in k-means

Author

Naldi, Murilo C. ; Fontana, André ; Campello, Ricardo J G B

Author_Institution

Comput. Sci. Dept., Univ. of Sao Paulo (USP) at Sao Carlos, Sao Carlos, Brazil

fYear

2009

fDate

Nov. 30 2009-Dec. 2 2009

Firstpage

1006

Lastpage

1013

Abstract

One of the most influential algorithms in data mining, k-means, is broadly used in practical tasks for its simplicity, computational efficiency and effectiveness in high dimensional problems. However, k-means has two major drawbacks, which are the need to choose the number of clusters, k, and the sensibility to the initial prototypes´ position. In this work, systematic, evolutionary and order heuristics used to suppress these drawbacks are compared. 27 variants of 4 algorithmic approaches are used to partition 324 synthetic data sets and the obtained results are compared.

Keywords

evolutionary computation; pattern clustering; data mining; evolutionary heuristics; k estimation; k-means; order heuristics; pattern clustering; Application software; Clustering algorithms; Computational efficiency; Computer science; Data mining; Evolutionary computation; Intelligent systems; Iterative algorithms; Partitioning algorithms; Prototypes; data mining; evolutionary computation;

fLanguage

English

Publisher

ieee

Conference_Titel

Intelligent Systems Design and Applications, 2009. ISDA '09. Ninth International Conference on

Conference_Location

Pisa

Print_ISBN

978-1-4244-4735-0

Electronic_ISBN

978-0-7695-3872-3

Type

conf

DOI

10.1109/ISDA.2009.78

Filename

5364434