Flexible independent component analysis

Author

Seungjin Choi ; Cichocki, Andrzej ; Amari, Shunichi

Author_Institution

Sch. of Electr. & Electron. Eng., Chungbuk Nat. Univ., South Korea

fYear

1998

fDate

31 Aug-2 Sep 1998

Firstpage

83

Lastpage

92

Abstract

We present a flexible independent component analysis (ICA) algorithm which can separate mixtures of sub- and super-Gaussian source signals with self-adaptive nonlinearities. A flexible ICA algorithm, in the framework of natural Riemannian gradient, is derived using the parametrized generalized Gaussian density model. The nonlinear function in the flexible ICA algorithm is self-adaptive and is controlled by Gaussian exponent. Computer simulation results confirm the validity and high performance of the proposed algorithm

Keywords

Gaussian distribution; adaptive signal detection; feedforward neural nets; learning (artificial intelligence); statistical analysis; Gaussian density model; Gaussian distribution; Gaussian source separation; Riemannian gradient; feedforward neural net; flexible independent component analysis; learning algorithm; nonlinear function; probability; signal detection; Application software; Biomedical computing; High performance computing; Image analysis; Independent component analysis; Information analysis; Information systems; Maximum likelihood estimation; Robustness; Vectors;

fLanguage

English

Publisher

ieee

Conference_Titel

Neural Networks for Signal Processing VIII, 1998. Proceedings of the 1998 IEEE Signal Processing Society Workshop

Conference_Location

Cambridge

ISSN

1089-3555

Print_ISBN

0-7803-5060-X

Type

conf

DOI

10.1109/NNSP.1998.710637

Filename

710637