ICA with sparse connections: Revisited

Kun Zhang*, Heng Peng, Laiwan Chan, Aapo Hyvarinen

*Corresponding author for this work

Research output: Chapter in book/report/conference proceedingConference proceedingpeer-review

16 Citations (Scopus)


When applying independent component analysis (ICA), sometimes we expect the connections between the observed mixtures and the recovered independent components (or the original sources) to be sparse, to make the interpretation easier or to reduce the random effect in the results. In this paper we propose two methods to tackle this problem. One is based on adaptive Lasso, which exploits the L1 penalty with data-adaptive weights. We show the relationship between this method and the classic information criteria such as BIC and AIC. The other is based on optimal brain surgeon, and we show how its stopping criterion is related to the information criteria. This method produces the solution path of the transformation matrix, with different number of zero entries. These methods involve low computational loads. Moreover, in each method, the parameter controlling the sparsity level of the transformation matrix has clear interpretations. By setting such parameters to certain values, the results of the proposed methods are consistent with those produced by classic information criteria.

Original languageEnglish
Title of host publicationIndependent Component Analysis and Signal Separation
Subtitle of host publication8th International Conference, ICA 2009, Paraty, Brazil, March 15-18, 2009, Proceedings
EditorsTülay Adali, Christian Jutten, João Marcos Travassos Romano, Allan Kardec Barros
Place of PublicationBerlin, Heidelberg
Number of pages8
ISBN (Electronic)9783642005992
ISBN (Print)9783642005985
Publication statusPublished - 25 Feb 2009
Event8th International Conference on Independent Component Analysis and Signal Separation, ICA 2009 - Paraty, Brazil
Duration: 15 Mar 200918 Mar 2009

Publication series

NameLecture Notes in Computer Science
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349
NameTheoretical Computer Science and General Issues
NameICA: International Conference on Independent Component Analysis and Signal Separation


Conference8th International Conference on Independent Component Analysis and Signal Separation, ICA 2009
Internet address

Scopus Subject Areas

  • Theoretical Computer Science
  • Computer Science(all)

User-Defined Keywords

  • Independent Component Analysis
  • Sparsity Level
  • Adaptive Lasso
  • Oracle Property


Dive into the research topics of 'ICA with sparse connections: Revisited'. Together they form a unique fingerprint.

Cite this