Blind source separation using block-coordinate relative Newton method

Alexander M. Bronstein, Michael M. Bronstein, Michael Zibulevsky*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

Presented here is a block-coordinate version of the relative Newton method, recently proposed for quasi-maximum likelihood blind source separation. Special structure of the Hessian matrix allows performing block-coordinate Newton descent efficiently. Simulations show that typically our method converges in near constant number of iterations (order of 10) independently of the problem size.

Original languageEnglish
Pages (from-to)1447-1459
Number of pages13
JournalSignal Processing
Volume84
Issue number8
DOIs
StatePublished - Aug 2004
Externally publishedYes

Keywords

  • Blind source separation
  • Block-coordinate optimization
  • Newton algorithm
  • Quasi-maximum likelihood

Fingerprint

Dive into the research topics of 'Blind source separation using block-coordinate relative Newton method'. Together they form a unique fingerprint.

Cite this