DocumentCode
394611
Title
Relative Newton method for signal separation
Author
Zibulevsky, Michael
Author_Institution
Dept. of Electr. Eng., Technion-Israel Inst. of Technol., Haifa, Israel
Volume
5
fYear
2003
fDate
6-10 April 2003
Abstract
The presented relative Newton method for quasi-maximum likelihood blind source separation significantly outperforms the natural gradient descent in batch mode. The structure of the corresponding Hessian matrix allows its fast inversion without assembling. Experiments with sparsely representable signals demonstrate super-efficient separation. More experiments with natural images are presented elsewhere (http://ie.technion.ac.il/∼mcib/).
Keywords
Hessian matrices; Newton method; blind source separation; optimisation; Hessian matrix; natural gradient descent; natural images; quasi-maximum likelihood blind source separation; relative Newton method; relative optimization; signal separation; sparse signals; Assembly; Blind source separation; Convergence; Gradient methods; Matrix converters; Minimization methods; Newton method; Probability density function; Source separation; Sparse matrices;
fLanguage
English
Publisher
ieee
Conference_Titel
Acoustics, Speech, and Signal Processing, 2003. Proceedings. (ICASSP '03). 2003 IEEE International Conference on
ISSN
1520-6149
Print_ISBN
0-7803-7663-3
Type
conf
DOI
10.1109/ICASSP.2003.1199860
Filename
1199860
Link To Document