# Recursive self-organizing map as a contractive iterative function system

Peter Tiňo, Igor Farkaš, Jort van Mourik

Research output: Chapter in Book/Report/Conference proceedingChapter

### Abstract

Recently, there has been a considerable research activity in extending topographic maps of vectorial data to more general data structures, such as sequences or trees. However, the representational capabilities and internal representations of the models are not well understood. We rigorously analyze a generalization of the Self-Organizing Map (SOM) for processing sequential data, Recursive SOM (RecSOM [1]), as a non-autonomous dynamical system consisting off a set of fixed input maps. We show that contractive fixed input maps are likely to produce Markovian organizations of receptive fields o the RecSOM map. We derive bounds on parameter $\beta$ (weighting the importance of importing past information when processing sequences) under which contractiveness of the fixed input maps is guaranteed.
Original language English Intelligent Data Engineering and Automated Learning - IDEAL 2005 Marcus Gallagher, James Hogan, Frederic Maire Berlin (DE) Springer 327-334 8 978-3-540-26972-4 https://doi.org/10.1007/11508069_43 Published - 20 Jun 2005

### Publication series

Name Lecture Notes in Computer Science Springer-Verlag 3578

### Fingerprint

Self organizing maps
Data structures
Dynamical systems
Processing

### Keywords

• topographic maps
• vectorial data
• contractive fixed input map
• Markovian organization
• receptive fields
• RecSOM map

### Cite this

Tiňo, P., Farkaš, I., & van Mourik, J. (2005). Recursive self-organizing map as a contractive iterative function system. In M. Gallagher, J. Hogan, & F. Maire (Eds.), Intelligent Data Engineering and Automated Learning - IDEAL 2005 (pp. 327-334). (Lecture Notes in Computer Science; Vol. 3578). Berlin (DE): Springer. https://doi.org/10.1007/11508069_43
Tiňo, Peter ; Farkaš, Igor ; van Mourik, Jort. / Recursive self-organizing map as a contractive iterative function system. Intelligent Data Engineering and Automated Learning - IDEAL 2005. editor / Marcus Gallagher ; James Hogan ; Frederic Maire. Berlin (DE) : Springer, 2005. pp. 327-334 (Lecture Notes in Computer Science).
@inbook{3d16f78b35dd4e8da29a93dcc5e5066f,
title = "Recursive self-organizing map as a contractive iterative function system",
abstract = "Recently, there has been a considerable research activity in extending topographic maps of vectorial data to more general data structures, such as sequences or trees. However, the representational capabilities and internal representations of the models are not well understood. We rigorously analyze a generalization of the Self-Organizing Map (SOM) for processing sequential data, Recursive SOM (RecSOM [1]), as a non-autonomous dynamical system consisting off a set of fixed input maps. We show that contractive fixed input maps are likely to produce Markovian organizations of receptive fields o the RecSOM map. We derive bounds on parameter $\beta$ (weighting the importance of importing past information when processing sequences) under which contractiveness of the fixed input maps is guaranteed.",
keywords = "topographic maps, vectorial data, contractive fixed input map, Markovian organization, receptive fields, RecSOM map",
author = "Peter Tiňo and Igor Farkaš and {van Mourik}, Jort",
year = "2005",
month = "6",
day = "20",
doi = "10.1007/11508069_43",
language = "English",
isbn = "978-3-540-26972-4",
series = "Lecture Notes in Computer Science",
publisher = "Springer",
pages = "327--334",
editor = "Marcus Gallagher and James Hogan and Frederic Maire",
booktitle = "Intelligent Data Engineering and Automated Learning - IDEAL 2005",

}

Tiňo, P, Farkaš, I & van Mourik, J 2005, Recursive self-organizing map as a contractive iterative function system. in M Gallagher, J Hogan & F Maire (eds), Intelligent Data Engineering and Automated Learning - IDEAL 2005. Lecture Notes in Computer Science, vol. 3578, Springer, Berlin (DE), pp. 327-334. https://doi.org/10.1007/11508069_43

Recursive self-organizing map as a contractive iterative function system. / Tiňo, Peter; Farkaš, Igor; van Mourik, Jort.

Intelligent Data Engineering and Automated Learning - IDEAL 2005. ed. / Marcus Gallagher; James Hogan; Frederic Maire. Berlin (DE) : Springer, 2005. p. 327-334 (Lecture Notes in Computer Science; Vol. 3578).

Research output: Chapter in Book/Report/Conference proceedingChapter

TY - CHAP

T1 - Recursive self-organizing map as a contractive iterative function system

AU - Tiňo, Peter

AU - Farkaš, Igor

AU - van Mourik, Jort

PY - 2005/6/20

Y1 - 2005/6/20

N2 - Recently, there has been a considerable research activity in extending topographic maps of vectorial data to more general data structures, such as sequences or trees. However, the representational capabilities and internal representations of the models are not well understood. We rigorously analyze a generalization of the Self-Organizing Map (SOM) for processing sequential data, Recursive SOM (RecSOM [1]), as a non-autonomous dynamical system consisting off a set of fixed input maps. We show that contractive fixed input maps are likely to produce Markovian organizations of receptive fields o the RecSOM map. We derive bounds on parameter $\beta$ (weighting the importance of importing past information when processing sequences) under which contractiveness of the fixed input maps is guaranteed.

AB - Recently, there has been a considerable research activity in extending topographic maps of vectorial data to more general data structures, such as sequences or trees. However, the representational capabilities and internal representations of the models are not well understood. We rigorously analyze a generalization of the Self-Organizing Map (SOM) for processing sequential data, Recursive SOM (RecSOM [1]), as a non-autonomous dynamical system consisting off a set of fixed input maps. We show that contractive fixed input maps are likely to produce Markovian organizations of receptive fields o the RecSOM map. We derive bounds on parameter $\beta$ (weighting the importance of importing past information when processing sequences) under which contractiveness of the fixed input maps is guaranteed.

KW - topographic maps

KW - vectorial data

KW - contractive fixed input map

KW - Markovian organization

KW - receptive fields

KW - RecSOM map

UR - http://www.scopus.com/inward/record.url?scp=26444435652&partnerID=8YFLogxK

UR - http://www.springer.com/computer/database+management+%26+information+retrieval/book/978-3-540-26972-4

U2 - 10.1007/11508069_43

DO - 10.1007/11508069_43

M3 - Chapter

SN - 978-3-540-26972-4

T3 - Lecture Notes in Computer Science

SP - 327

EP - 334

BT - Intelligent Data Engineering and Automated Learning - IDEAL 2005

A2 - Gallagher, Marcus

A2 - Hogan, James

A2 - Maire, Frederic

PB - Springer

CY - Berlin (DE)

ER -

Tiňo P, Farkaš I, van Mourik J. Recursive self-organizing map as a contractive iterative function system. In Gallagher M, Hogan J, Maire F, editors, Intelligent Data Engineering and Automated Learning - IDEAL 2005. Berlin (DE): Springer. 2005. p. 327-334. (Lecture Notes in Computer Science). https://doi.org/10.1007/11508069_43