Title :
A New Non-Orthogonal Joint Diagonalization Algorithm with Application in ICA and BSS
Author :
Fuxiang, Wang ; Chongkan, Liu ; Jun, Zhang
Author_Institution :
Sch. of Electron. & Inf. Eng., Beihang Univ., Beijing
Abstract :
Joint diagonalization is an important family of methods for ICA (independent component analysis) and BSS (blind source separation). In this paper, we present a new approach to joint diagonalization of a set of 2x2 real symmetric matrices with a general (not necessarily orthogonal) matrix. It is a non-iterative algorithm in which the joint diagonalization problem can be solved by a special eigenvector problem. Some blind source separation (BSS) simulations results demonstrate that the algorithm has a substantial improvement in the separating performance.
Keywords :
blind source separation; eigenvalues and eigenfunctions; independent component analysis; matrix algebra; BSS; ICA; blind source separation; eigenvector problem; independent component analysis; noniterative algorithm; nonorthogonal joint diagonalization algorithm; symmetric matrices; Additive noise; Blind source separation; Computational Intelligence Society; Cost function; Data models; Independent component analysis; Source separation; Statistics; Symmetric matrices; Systems engineering and theory; Blind source separation; independent component analysis; joint diagonalization;
Conference_Titel :
Computational Engineering in Systems Applications, IMACS Multiconference on
Conference_Location :
Beijing
Print_ISBN :
7-302-13922-9
Electronic_ISBN :
7-900718-14-1
DOI :
10.1109/CESA.2006.4281629
