Robust Topology Representing Networks

Michaël Aupetit

Robust Topology Representing Networks

Michaël Aupetit^*

^*Corresponding author for this work

Commissariat à l’énergie atomique et aux énergies alternatives

Research output: Contribution to conference › Paper › peer-review

Abstract

Martinetz and Schulten proposed the use of a Competitive Hebbian Learning (CHL) rule to build Topology Representing Networks. From a set of units and a data distribution, a link is created between the first and second closest units to each datum, creating a graph which preserves the topology of the data set. However, one has to deal with finite data distributions generally corrupted with noise, for which CHL may be unefficient. We propose a more robust approach to create a topology representing graph, by considering the density of the data distribution.

Original language	English
Pages	45-50
Number of pages	6
Publication status	Published - 2003
Externally published	Yes
Event	11th European Symposium on Artificial Neural Networks, ESANN 2003 - Bruges, Belgium Duration: 23 Apr 2003 → 25 Apr 2003

Conference

Conference	11th European Symposium on Artificial Neural Networks, ESANN 2003
Country/Territory	Belgium
City	Bruges
Period	23/04/03 → 25/04/03

Cite this

@conference{6471784c23004a79ae8856ed66f4e52c,

title = "Robust Topology Representing Networks",

abstract = "Martinetz and Schulten proposed the use of a Competitive Hebbian Learning (CHL) rule to build Topology Representing Networks. From a set of units and a data distribution, a link is created between the first and second closest units to each datum, creating a graph which preserves the topology of the data set. However, one has to deal with finite data distributions generally corrupted with noise, for which CHL may be unefficient. We propose a more robust approach to create a topology representing graph, by considering the density of the data distribution.",

author = "Micha{\"e}l Aupetit",

year = "2003",

language = "English",

pages = "45--50",

}

TY - CONF

T1 - Robust Topology Representing Networks

AU - Aupetit, Michaël

PY - 2003

Y1 - 2003

N2 - Martinetz and Schulten proposed the use of a Competitive Hebbian Learning (CHL) rule to build Topology Representing Networks. From a set of units and a data distribution, a link is created between the first and second closest units to each datum, creating a graph which preserves the topology of the data set. However, one has to deal with finite data distributions generally corrupted with noise, for which CHL may be unefficient. We propose a more robust approach to create a topology representing graph, by considering the density of the data distribution.

AB - Martinetz and Schulten proposed the use of a Competitive Hebbian Learning (CHL) rule to build Topology Representing Networks. From a set of units and a data distribution, a link is created between the first and second closest units to each datum, creating a graph which preserves the topology of the data set. However, one has to deal with finite data distributions generally corrupted with noise, for which CHL may be unefficient. We propose a more robust approach to create a topology representing graph, by considering the density of the data distribution.

UR - https://www.scopus.com/pages/publications/105034858624

M3 - Paper

AN - SCOPUS:105034858624

SP - 45

EP - 50

T2 - 11th European Symposium on Artificial Neural Networks, ESANN 2003

Y2 - 23 April 2003 through 25 April 2003

ER -

Robust Topology Representing Networks

Abstract

Conference

Other files and links

Fingerprint

Cite this