ISOMAP based metrics for clustering

Fuente: Redalyc
Enregistré dans:
Détails bibliographiques
Auteur principal: Ariel E. Bayá
Format: Artículo científico
Langue:en
Publié: Asociación Española para la Inteligencia Artificial 2008
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1876452755340525568
author Ariel E. Bayá
author_facet Ariel E. Bayá
contents ISOMAP based metrics for clustering Ariel E. Bayá Pablo M. Granitto Ingeniería Non ISOMAP Clustering linear Metric Affinity Propagation Many successful clustering techniques fail to handle data with a manifold structure, i.e. data that is notshaped in the form of compact point clouds, forming arbitrary shapes or paths through a high-dimensionalspace. In this paper we present a new method to evaluate distances in such spaces that naturally extend theapplication of many clustering algorithms to these cases. Our algorithm has two stages. Following ISOMAP,it searches for sets of locally-uniform manifolds, which could be disjoint. These manifolds are then connectedusing two slightly different strategies. We compare these strategies between them and with a state of theart algorithm using three artificial problems, obtaining encouraging result. Both new metrics allow diversealgorithms to easily find clusters of arbitrary shape 2008 artículo científico 1137-3601 https://www.redalyc.org/articulo.oa?id=92503703 en http://www.redalyc.org/revista.oa?id=925 Inteligencia Artificial. Revista Iberoamericana de Inteligencia Artificial application/pdf Asociación Española para la Inteligencia Artificial Inteligencia Artificial. Revista Iberoamericana de Inteligencia Artificial (España) Num.37 Vol.12
format Artículo científico
id redalyc_92503703
institution Redalyc
language en
publishDate 2008
publisher Asociación Española para la Inteligencia Artificial
spellingShingle ISOMAP based metrics for clustering
Ariel E. Bayá
Ingeniería
Non
ISOMAP
Clustering
linear Metric
Affinity Propagation
ISOMAP based metrics for clustering Ariel E. Bayá Pablo M. Granitto Ingeniería Non ISOMAP Clustering linear Metric Affinity Propagation Many successful clustering techniques fail to handle data with a manifold structure, i.e. data that is notshaped in the form of compact point clouds, forming arbitrary shapes or paths through a high-dimensionalspace. In this paper we present a new method to evaluate distances in such spaces that naturally extend theapplication of many clustering algorithms to these cases. Our algorithm has two stages. Following ISOMAP,it searches for sets of locally-uniform manifolds, which could be disjoint. These manifolds are then connectedusing two slightly different strategies. We compare these strategies between them and with a state of theart algorithm using three artificial problems, obtaining encouraging result. Both new metrics allow diversealgorithms to easily find clusters of arbitrary shape 2008 artículo científico 1137-3601 https://www.redalyc.org/articulo.oa?id=92503703 en http://www.redalyc.org/revista.oa?id=925 Inteligencia Artificial. Revista Iberoamericana de Inteligencia Artificial application/pdf Asociación Española para la Inteligencia Artificial Inteligencia Artificial. Revista Iberoamericana de Inteligencia Artificial (España) Num.37 Vol.12
title ISOMAP based metrics for clustering
topic Ingeniería
Non
ISOMAP
Clustering
linear Metric
Affinity Propagation
url https://www.redalyc.org/articulo.oa?id=92503703