Learning topology of a labeled data set with the supervised generative Gaussian graph

Pierre Gaillard, Michael Aupetit, Gérard Govaert

Research output: Contribution to journalArticle

10 Citations (Scopus)

Abstract

Extracting the topology of a set of a labeled data is expected to provide important information in order to analyze the data or to design a better decision system. In this work, we propose to extend the generative Gaussian graph to supervised learning in order to extract the topology of labeled data sets. The graph obtained learns the intra-class and inter-class connectedness and also the manifold-overlapping of the different classes. We propose a way to vizualize these topological features. We apply it to analyze the well-known Iris database and the three-phase pipe flow database.

Original languageEnglish
Pages (from-to)1283-1299
Number of pages17
JournalNeurocomputing
Volume71
Issue number7-9
DOIs
Publication statusPublished - Mar 2008
Externally publishedYes

Fingerprint

Topology
Learning
Databases
Supervised learning
Pipe flow
Iris
Datasets

Keywords

  • Delaunay graph
  • EM algorithm
  • Gabriel graph
  • Mixture models
  • Supervised topology learning
  • Topology representing graph

ASJC Scopus subject areas

  • Artificial Intelligence
  • Cellular and Molecular Neuroscience

Cite this

Learning topology of a labeled data set with the supervised generative Gaussian graph. / Gaillard, Pierre; Aupetit, Michael; Govaert, Gérard.

In: Neurocomputing, Vol. 71, No. 7-9, 03.2008, p. 1283-1299.

Research output: Contribution to journalArticle

Gaillard, Pierre ; Aupetit, Michael ; Govaert, Gérard. / Learning topology of a labeled data set with the supervised generative Gaussian graph. In: Neurocomputing. 2008 ; Vol. 71, No. 7-9. pp. 1283-1299.
@article{dd56019528644580b9a5e91deaa02768,
title = "Learning topology of a labeled data set with the supervised generative Gaussian graph",
abstract = "Extracting the topology of a set of a labeled data is expected to provide important information in order to analyze the data or to design a better decision system. In this work, we propose to extend the generative Gaussian graph to supervised learning in order to extract the topology of labeled data sets. The graph obtained learns the intra-class and inter-class connectedness and also the manifold-overlapping of the different classes. We propose a way to vizualize these topological features. We apply it to analyze the well-known Iris database and the three-phase pipe flow database.",
keywords = "Delaunay graph, EM algorithm, Gabriel graph, Mixture models, Supervised topology learning, Topology representing graph",
author = "Pierre Gaillard and Michael Aupetit and G{\'e}rard Govaert",
year = "2008",
month = "3",
doi = "10.1016/j.neucom.2007.12.028",
language = "English",
volume = "71",
pages = "1283--1299",
journal = "Neurocomputing",
issn = "0925-2312",
publisher = "Elsevier",
number = "7-9",

}

TY - JOUR

T1 - Learning topology of a labeled data set with the supervised generative Gaussian graph

AU - Gaillard, Pierre

AU - Aupetit, Michael

AU - Govaert, Gérard

PY - 2008/3

Y1 - 2008/3

N2 - Extracting the topology of a set of a labeled data is expected to provide important information in order to analyze the data or to design a better decision system. In this work, we propose to extend the generative Gaussian graph to supervised learning in order to extract the topology of labeled data sets. The graph obtained learns the intra-class and inter-class connectedness and also the manifold-overlapping of the different classes. We propose a way to vizualize these topological features. We apply it to analyze the well-known Iris database and the three-phase pipe flow database.

AB - Extracting the topology of a set of a labeled data is expected to provide important information in order to analyze the data or to design a better decision system. In this work, we propose to extend the generative Gaussian graph to supervised learning in order to extract the topology of labeled data sets. The graph obtained learns the intra-class and inter-class connectedness and also the manifold-overlapping of the different classes. We propose a way to vizualize these topological features. We apply it to analyze the well-known Iris database and the three-phase pipe flow database.

KW - Delaunay graph

KW - EM algorithm

KW - Gabriel graph

KW - Mixture models

KW - Supervised topology learning

KW - Topology representing graph

UR - http://www.scopus.com/inward/record.url?scp=40649110460&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=40649110460&partnerID=8YFLogxK

U2 - 10.1016/j.neucom.2007.12.028

DO - 10.1016/j.neucom.2007.12.028

M3 - Article

AN - SCOPUS:40649110460

VL - 71

SP - 1283

EP - 1299

JO - Neurocomputing

JF - Neurocomputing

SN - 0925-2312

IS - 7-9

ER -