A simple overcomplete ICA algorithm by non-orthogonal pair optimizations

Author

Matsuda, Yoshitatsu ; Yamaguchi, Kazunori

Author_Institution

Dept. of Integrated Inf. Technol., Aoyama Gakuin Univ., Sagamihara, Japan

fYear

2009

fDate

14-19 June 2009

Firstpage

2027

Lastpage

2031

Abstract

Overcomplete ICA is a method for solving blind source separation problems if the number of observed signals is less than that of source ones. In this paper, we propose an overcomplete ICA algorithm based on a simple contrast function which is defined as the sum of the covariances of the squares of signals over all the pairs. By applying non-orthogonal pair optimizations to the function, a simple ICA algorithm is derived. Theoretical analysis and numerical experiments suggest the validity of the proposed algorithm.

Keywords

Jacobian matrices; blind source separation; independent component analysis; optimisation; blind source separation problem; convariance sum; independent component analysis; nonorthogonal pair optimization; simple contrast function; simple overcomplete ICA; Algorithm design and analysis; Blind source separation; Independent component analysis; Information technology; Jacobian matrices; Neural networks; Newton method; Optimization methods;

fLanguage

English

Publisher

ieee

Conference_Titel

Neural Networks, 2009. IJCNN 2009. International Joint Conference on

Conference_Location

Atlanta, GA

ISSN

1098-7576

Print_ISBN

978-1-4244-3548-7

Electronic_ISBN

1098-7576

Type

conf

DOI

10.1109/IJCNN.2009.5178945

Filename

5178945