Robust ICA for super-Gaussian sources

Frank C. Meinecke; Stefan Harmeling; Klaus Robert Müller

doi:10.1007/978-3-540-30110-3_28

Robust ICA for super-Gaussian sources

Frank C. Meinecke, Stefan Harmeling, Klaus Robert Müller

Research output: Chapter in Book/Report/Conference proceeding › Chapter

6 Citations (Scopus)

Abstract

Most ICA algorithms are sensitive to outliers. Instead of robustifying existing algorithms by outlier rejection techniques, we show how a simple outlier index can be used directly to solve the ICA problem for super-Gaussian source signals. This ICA method is outlier-robust by construction and can be used for standard ICA as well as for overcomplete ICA (i.e. more source signals than observed signals (mixtures)).

Original language	English
Title of host publication	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Editors	Carlos G. Puntonet, Alberto Prieto
Publisher	Springer Verlag
Pages	217-224
Number of pages	8
ISBN (Electronic)	3540230564, 9783540230564
DOIs	https://doi.org/10.1007/978-3-540-30110-3_28
Publication status	Published - 2004
Externally published	Yes

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	3195
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

ASJC Scopus subject areas

Theoretical Computer Science
Computer Science(all)

Access to Document

10.1007/978-3-540-30110-3_28

Fingerprint Dive into the research topics of 'Robust ICA for super-Gaussian sources'. Together they form a unique fingerprint.

Cite this

Meinecke, F. C., Harmeling, S., & Müller, K. R. (2004). Robust ICA for super-Gaussian sources. In C. G. Puntonet, & A. Prieto (Eds.), Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (pp. 217-224). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 3195). Springer Verlag. https://doi.org/10.1007/978-3-540-30110-3_28

Robust ICA for super-Gaussian sources. / Meinecke, Frank C.; Harmeling, Stefan; Müller, Klaus Robert.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). ed. / Carlos G. Puntonet; Alberto Prieto. Springer Verlag, 2004. p. 217-224 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 3195).

Research output: Chapter in Book/Report/Conference proceeding › Chapter

Meinecke, FC, Harmeling, S & Müller, KR 2004, Robust ICA for super-Gaussian sources. in CG Puntonet & A Prieto (eds), Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 3195, Springer Verlag, pp. 217-224. https://doi.org/10.1007/978-3-540-30110-3_28

Meinecke FC, Harmeling S, Müller KR. Robust ICA for super-Gaussian sources. In Puntonet CG, Prieto A, editors, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Springer Verlag. 2004. p. 217-224. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-540-30110-3_28

Meinecke, Frank C. ; Harmeling, Stefan ; Müller, Klaus Robert. / Robust ICA for super-Gaussian sources. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). editor / Carlos G. Puntonet ; Alberto Prieto. Springer Verlag, 2004. pp. 217-224 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inbook{59a7624cce5640b8b80f4ffe524a42e5,

title = "Robust ICA for super-Gaussian sources",

abstract = "Most ICA algorithms are sensitive to outliers. Instead of robustifying existing algorithms by outlier rejection techniques, we show how a simple outlier index can be used directly to solve the ICA problem for super-Gaussian source signals. This ICA method is outlier-robust by construction and can be used for standard ICA as well as for overcomplete ICA (i.e. more source signals than observed signals (mixtures)).",

author = "Meinecke, {Frank C.} and Stefan Harmeling and M{\"u}ller, {Klaus Robert}",

year = "2004",

doi = "10.1007/978-3-540-30110-3_28",

language = "English",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "217--224",

editor = "Puntonet, {Carlos G.} and Alberto Prieto",

booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

}

TY - CHAP

T1 - Robust ICA for super-Gaussian sources

AU - Meinecke, Frank C.

AU - Harmeling, Stefan

AU - Müller, Klaus Robert

PY - 2004

Y1 - 2004

N2 - Most ICA algorithms are sensitive to outliers. Instead of robustifying existing algorithms by outlier rejection techniques, we show how a simple outlier index can be used directly to solve the ICA problem for super-Gaussian source signals. This ICA method is outlier-robust by construction and can be used for standard ICA as well as for overcomplete ICA (i.e. more source signals than observed signals (mixtures)).

AB - Most ICA algorithms are sensitive to outliers. Instead of robustifying existing algorithms by outlier rejection techniques, we show how a simple outlier index can be used directly to solve the ICA problem for super-Gaussian source signals. This ICA method is outlier-robust by construction and can be used for standard ICA as well as for overcomplete ICA (i.e. more source signals than observed signals (mixtures)).

UR - http://www.scopus.com/inward/record.url?scp=35048843771&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=35048843771&partnerID=8YFLogxK

U2 - 10.1007/978-3-540-30110-3_28

DO - 10.1007/978-3-540-30110-3_28

M3 - Chapter

AN - SCOPUS:35048843771

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 217

EP - 224

BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

A2 - Puntonet, Carlos G.

A2 - Prieto, Alberto

PB - Springer Verlag

ER -