Title :
Redundancy reduction: a new criterion for blind separation of signals
Author :
Liu, Ju ; Nie, Kaibao ; He, Zhenya
Author_Institution :
Dept. of Electron. Eng., Shandong Univ., Jinan, China
Abstract :
In this paper, we proposed an information theory criterion for blind separation of signals. Based on the criterion, we developed an algorithm to maximize transferred information. A Hebb term is introduced in the algorithm to approximate the derivative of information loss in information transmission. The mutual information of outputs is minimized to make the outputs mutually independent. Computer simulations with generated mixtures show the validity of the proposed algorithm
Keywords :
information theory; redundancy; signal processing; Hebb term; blind separation; information loss; information theory criterion; information transmission; redundancy reduction; signals separation; Entropy; Feedforward neural networks; Feedforward systems; Independent component analysis; Measurement uncertainty; Mutual information; Neural networks; Probability density function; Propagation losses; Random variables;
Conference_Titel :
Circuits and Systems, 2000. IEEE APCCAS 2000. The 2000 IEEE Asia-Pacific Conference on
Conference_Location :
Tianjin
Print_ISBN :
0-7803-6253-5
DOI :
10.1109/APCCAS.2000.913569
