Loss function for blind source separation-minimum entropy criterion and its generalized anti-Hebbian rules

Hsiao-Chun Wu, J.C. Principe, J.G. Harris, Jui-Kuo Juan
IJCNN'99. International Joint Conference on Neural Networks. Proceedings (Cat. No.99CH36339)  
Blind source separation has been intriguing many scientists. In adaptive signal processing, LMS (kast-mean squared) algorithm has long been used in signal enhancement and noise cancellation but it cannot ovexome the d$jiculty caused by the signal leakage into the reference input. Hence we have to explore more general statistical properties about the observed signals. This view corresponds to a statistical modeling of the signals using statistical measures such as a loss function, which is
more » ... ent from the mutual information. This paper will propose a new loss function based on generalized Gaussian distribution family and derive nav simple adaptive learning rules. Our separator based on the new generalized "anti-Hebbian rules" is also justified by the simulation on both artificial and real data with good pelfonnance.
doi:10.1109/ijcnn.1999.831074 dblp:conf/ijcnn/WuPHJ99 fatcat:27ja7mafyva43cent33jelc45m