Independent subspace analysis can cope with the 'curse of dimensionality'

Zoltán Szabó, A. Lőrincz

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

We search for hidden independent components, in particular we consider the independent subspace analysis (ISA) task. Earlier ISA procedures assume that the dimensions of the components are known. Here we show a method that enables the non-combinatorial estimation of the components. We make use of a decomposition principle called the ISA separation theorem. According to this separation theorem the ISA task can be reduced to the independent component analysis (ICA) task that assumes one-dimensional components and then to a grouping procedure that collects the respective non-independent elements into independent groups. We show that non-combinatorial grouping is feasible by means of the non-linear f-correlation matrices between the estimated components.

Original languageEnglish
Pages (from-to)213-221
Number of pages9
JournalActa Cybernetica
Volume18
Issue number2
Publication statusPublished - 2007

Fingerprint

Curse of Dimensionality
Subspace
Independent component analysis
Separation Theorem
Decomposition
Grouping
Correlation Matrix
Independent Component Analysis
Curse of dimensionality
Decompose

Keywords

  • Independent subspace analysis
  • Non-combinatorial solution

ASJC Scopus subject areas

  • Hardware and Architecture
  • Software
  • Computational Theory and Mathematics
  • Theoretical Computer Science

Cite this

Independent subspace analysis can cope with the 'curse of dimensionality'. / Szabó, Zoltán; Lőrincz, A.

In: Acta Cybernetica, Vol. 18, No. 2, 2007, p. 213-221.

Research output: Contribution to journalArticle

@article{8a72a1f3b910407cae588f0276202387,
title = "Independent subspace analysis can cope with the 'curse of dimensionality'",
abstract = "We search for hidden independent components, in particular we consider the independent subspace analysis (ISA) task. Earlier ISA procedures assume that the dimensions of the components are known. Here we show a method that enables the non-combinatorial estimation of the components. We make use of a decomposition principle called the ISA separation theorem. According to this separation theorem the ISA task can be reduced to the independent component analysis (ICA) task that assumes one-dimensional components and then to a grouping procedure that collects the respective non-independent elements into independent groups. We show that non-combinatorial grouping is feasible by means of the non-linear f-correlation matrices between the estimated components.",
keywords = "Independent subspace analysis, Non-combinatorial solution",
author = "Zolt{\'a}n Szab{\'o} and A. Lőrincz",
year = "2007",
language = "English",
volume = "18",
pages = "213--221",
journal = "Acta Cybernetica",
issn = "0324-721X",
publisher = "University of Szeged",
number = "2",

}

TY - JOUR

T1 - Independent subspace analysis can cope with the 'curse of dimensionality'

AU - Szabó, Zoltán

AU - Lőrincz, A.

PY - 2007

Y1 - 2007

N2 - We search for hidden independent components, in particular we consider the independent subspace analysis (ISA) task. Earlier ISA procedures assume that the dimensions of the components are known. Here we show a method that enables the non-combinatorial estimation of the components. We make use of a decomposition principle called the ISA separation theorem. According to this separation theorem the ISA task can be reduced to the independent component analysis (ICA) task that assumes one-dimensional components and then to a grouping procedure that collects the respective non-independent elements into independent groups. We show that non-combinatorial grouping is feasible by means of the non-linear f-correlation matrices between the estimated components.

AB - We search for hidden independent components, in particular we consider the independent subspace analysis (ISA) task. Earlier ISA procedures assume that the dimensions of the components are known. Here we show a method that enables the non-combinatorial estimation of the components. We make use of a decomposition principle called the ISA separation theorem. According to this separation theorem the ISA task can be reduced to the independent component analysis (ICA) task that assumes one-dimensional components and then to a grouping procedure that collects the respective non-independent elements into independent groups. We show that non-combinatorial grouping is feasible by means of the non-linear f-correlation matrices between the estimated components.

KW - Independent subspace analysis

KW - Non-combinatorial solution

UR - http://www.scopus.com/inward/record.url?scp=38049113161&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=38049113161&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:38049113161

VL - 18

SP - 213

EP - 221

JO - Acta Cybernetica

JF - Acta Cybernetica

SN - 0324-721X

IS - 2

ER -