Title :
Dependence of Bivariate Signals in ICA Problem
Author :
Changguo, Li ; Wenli, Suo
Author_Institution :
Dept. of Basic Sci., Mil. Transp. Univ., Tianjin, China
Abstract :
A new dependence measure of random variables is proposed as the contrast function of two dimensional ICA. By simulted annealing algorithm, variety of mixture of signals including subgaussian and supergaussian can be separated.
Keywords :
blind source separation; independent component analysis; simulated annealing; bivariate signal dependence; simulated annealing algorithm; subGaussian signal; superGaussian signal; two dimensional ICA problem; Approximation methods; Correlation; Covariance matrix; Gaussian distribution; Independent component analysis; Mutual information; Random variables; Chi-square statistic; Correlation; Dependence;
Conference_Titel :
Intelligent Systems (GCIS), 2010 Second WRI Global Congress on
Conference_Location :
Wuhan
Print_ISBN :
978-1-4244-9247-3
DOI :
10.1109/GCIS.2010.270
