New Riemannian metrics for speeding-up the convergence of over- And underdetermined ICA

Stefano Squartini; Francesco Piazza; Fabian J. Theis

New Riemannian metrics for speeding-up the convergence of over- And underdetermined ICA

Stefano Squartini, Francesco Piazza, Fabian J. Theis

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

4 Scopus citations

Abstract

In this paper some alternative Riemannian metrics are defined on the parameter space of non-square matrices, corresponding to various translations defined therein. Such metrics allow the authors to derive novel learning rules for two ICA based algorithms for over-determined blind source separation (BSS), which tries to separate less sources from more sensors. Computer simulations show a significant improvement of the convergence speed when second-order translations are employed in contrast to their first-order counterparts, extending known results for complete BSS.

Original language	English
Title of host publication	ISCAS 2006
Subtitle of host publication	2006 IEEE International Symposium on Circuits and Systems, Proceedings
Pages	4555-4558
Number of pages	4
State	Published - 2006
Externally published	Yes
Event	ISCAS 2006: 2006 IEEE International Symposium on Circuits and Systems - Kos, Greece Duration: 21 May 2006 → 24 May 2006

Publication series

Name	Proceedings - IEEE International Symposium on Circuits and Systems
ISSN (Print)	0271-4310

Conference

Conference	ISCAS 2006: 2006 IEEE International Symposium on Circuits and Systems
Country/Territory	Greece
City	Kos
Period	21/05/06 → 24/05/06

Cite this

@inproceedings{122b5ad89ea848d9ba24b716b2986765,

title = "New Riemannian metrics for speeding-up the convergence of over- And underdetermined ICA",

abstract = "In this paper some alternative Riemannian metrics are defined on the parameter space of non-square matrices, corresponding to various translations defined therein. Such metrics allow the authors to derive novel learning rules for two ICA based algorithms for over-determined blind source separation (BSS), which tries to separate less sources from more sensors. Computer simulations show a significant improvement of the convergence speed when second-order translations are employed in contrast to their first-order counterparts, extending known results for complete BSS.",

author = "Stefano Squartini and Francesco Piazza and Theis, {Fabian J.}",

year = "2006",

language = "English",

isbn = "0780393902",

series = "Proceedings - IEEE International Symposium on Circuits and Systems",

pages = "4555--4558",

booktitle = "ISCAS 2006",

note = "ISCAS 2006: 2006 IEEE International Symposium on Circuits and Systems ; Conference date: 21-05-2006 Through 24-05-2006",

}

Squartini, S, Piazza, F & Theis, FJ 2006, New Riemannian metrics for speeding-up the convergence of over- And underdetermined ICA. in ISCAS 2006: 2006 IEEE International Symposium on Circuits and Systems, Proceedings., 1693643, Proceedings - IEEE International Symposium on Circuits and Systems, pp. 4555-4558, ISCAS 2006: 2006 IEEE International Symposium on Circuits and Systems, Kos, Greece, 21/05/06.

New Riemannian metrics for speeding-up the convergence of over- And underdetermined ICA. / Squartini, Stefano; Piazza, Francesco; Theis, Fabian J.
ISCAS 2006: 2006 IEEE International Symposium on Circuits and Systems, Proceedings. 2006. p. 4555-4558 1693643 (Proceedings - IEEE International Symposium on Circuits and Systems).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - New Riemannian metrics for speeding-up the convergence of over- And underdetermined ICA

AU - Squartini, Stefano

AU - Piazza, Francesco

AU - Theis, Fabian J.

PY - 2006

Y1 - 2006

N2 - In this paper some alternative Riemannian metrics are defined on the parameter space of non-square matrices, corresponding to various translations defined therein. Such metrics allow the authors to derive novel learning rules for two ICA based algorithms for over-determined blind source separation (BSS), which tries to separate less sources from more sensors. Computer simulations show a significant improvement of the convergence speed when second-order translations are employed in contrast to their first-order counterparts, extending known results for complete BSS.

AB - In this paper some alternative Riemannian metrics are defined on the parameter space of non-square matrices, corresponding to various translations defined therein. Such metrics allow the authors to derive novel learning rules for two ICA based algorithms for over-determined blind source separation (BSS), which tries to separate less sources from more sensors. Computer simulations show a significant improvement of the convergence speed when second-order translations are employed in contrast to their first-order counterparts, extending known results for complete BSS.

UR - http://www.scopus.com/inward/record.url?scp=34547248473&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:34547248473

SN - 0780393902

SN - 9780780393905

T3 - Proceedings - IEEE International Symposium on Circuits and Systems

SP - 4555

EP - 4558

BT - ISCAS 2006

T2 - ISCAS 2006: 2006 IEEE International Symposium on Circuits and Systems

Y2 - 21 May 2006 through 24 May 2006

ER -

New Riemannian metrics for speeding-up the convergence of over- And underdetermined ICA

Abstract

Publication series

Conference

Other files and links

Fingerprint

Cite this