Article (Scientific journals)
The influence function of the TCLUST robust clustering procedure
Ruwet, Christel; García-Escudero, Luis Angel; Gordaliza, Alfonso et al.
2012In Advances in Data Analysis and Classification, 6 (2), p. 107-130
Peer Reviewed verified by ORBi
 

Files


Full Text
The_IF_of_the_tclust_procedure_TR.pdf
Author postprint (357.64 kB)
Request a copy

The original publication is available at www.springerlink.com


All documents in ORBi are protected by a user license.

Send to



Details



Keywords :
Heterogeneous Clustering; Influence function; Robustness; Trimmin
Abstract :
[en] The TCLUST procedure performs robust clustering with the aim of finding clusters with different scatter structures and proportions. An Eigenvalue Ratio constraint is considered by TCLUST in order to avoid finding spurious clusters. In order to guarantee the robustness of the method against the presence of outliers and background noise, the method allows for trimming of a given proportion of observations self determined by the data. This article studies robustness properties of the TCLUST procedure by means of the influence function, obtaining a robustness behavior close to that of the trimmed k-means.
Disciplines :
Mathematics
Author, co-author :
Ruwet, Christel ;  Université de Liège - ULiège > Département de mathématique > Statistique mathématique
García-Escudero, Luis Angel;  Universidad deValladolid - UVa
Gordaliza, Alfonso;  Universidad deValladolid - UVa
Mayo-Iscar, Agustin;  Universidad deValladolid - UVa
Language :
English
Title :
The influence function of the TCLUST robust clustering procedure
Publication date :
2012
Journal title :
Advances in Data Analysis and Classification
ISSN :
1862-5347
eISSN :
1862-5355
Publisher :
Springer, Germany
Volume :
6
Issue :
2
Pages :
107-130
Peer reviewed :
Peer Reviewed verified by ORBi
Funders :
Spanish Ministerio de Ciencia e Innovación
FWB - Fédération Wallonie-Bruxelles [BE]
Available on ORBi :
since 15 June 2011

Statistics


Number of views
85 (8 by ULiège)
Number of downloads
9 (0 by ULiège)

Scopus citations®
 
10
Scopus citations®
without self-citations
7
OpenCitations
 
7

Bibliography


Similar publications



Contact ORBi