|
Title:
|
ON APPLICATION OF PROBABILITY METRICS IN THE CLUSTER STABILITY PROBLEM |
|
Author(s):
|
Z. Volkovich , Z. Barzily |
|
ISBN:
|
978-972-8924-40-9 |
|
Editors:
|
Jörg Roth, Jairo Gutiérrez and Ajith P. Abraham (series editors: Piet Kommers, Pedro Isaías and Nian-Shing Chen) |
|
Year:
|
2007 |
|
Edition:
|
Single |
|
Keywords:
|
Clustering, Cluster Validation, Probability Metrics |
|
Type:
|
Full Paper |
|
First Page:
|
41 |
|
Last Page:
|
49 |
|
Language:
|
English |
|
Cover:
|
|
|
Full Contents:
|
click to dowload 
|
|
Paper Abstract:
|
In this paper we propose a general approach for the study of the cluster validation problem. Our method suggests the
application of probability metrics between random variables. Such a methodology produces a new point of view for the
investigation of the referenced problem and allows interpreting several existing methods from the same general attitude.
Our concept suggests that partitions obtained by a cluster algorithm rerunning can be considered as realizations of a unique
random variable such that the steadiest random variable infers the true number of clusters. We offer to measure this stability
by means of probability distances within the observed realization. |
|
|
|
|
|
|
