Post-nonlinear underdetermined ICA by bayesian statistics

Chen Wei*, Li Chin Khor, Wai Lok Woo, Satnam Singh Dlay

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

2 Citations (Scopus)


The problem of nonlinear signal separation and underdetermined signal separation has received increasing attention in the research of blind signal separation. Few of them can solve the situation where nonlinear and underdetermined characteristics exist simultaneously. In this paper, a new learning algorithm based on Bayesian statistics is proposed to solve the problem of the blind separation of nonlinear and underdetermined mixtures. This paper addresses the Blind Signal Separation (BSS) of post-nonlinearly mixed signals where the number of observations is less than the number of sources. Formal derivation shows that the source signals, mixing matrix and nonlinear functions can be estimated through an iterative technique based on alternate optimization. Simulations have been carried out to demonstrate the effectiveness of the proposed algorithm in separating signals under nonlinear and underdetermined conditions.

Original languageEnglish
Title of host publicationIndependent Component Analysis and Blind Signal Separation - 6th International Conference, ICA 2006, Proceedings
Number of pages8
ISBN (Electronic)9783540326311
ISBN (Print)9783540326304
Publication statusPublished - 2006
Event6th International Conference on Independent Component Analysis and Blind Signal Separation, ICA 2006 - Charleston, SC, United States
Duration: 5 Mar 20068 Mar 2006

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume3889 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Conference6th International Conference on Independent Component Analysis and Blind Signal Separation, ICA 2006
Country/TerritoryUnited States
CityCharleston, SC


Dive into the research topics of 'Post-nonlinear underdetermined ICA by bayesian statistics'. Together they form a unique fingerprint.

Cite this