Title :
On blind separation of complex-valued sources by extended Hebbian learning
Author_Institution :
Dept. of Ind. Eng., Perugia Univ., Italy
Abstract :
The aim of this letter is to present a nonlinear extension to Sanger´s (1989) generalized Hebbian learning algorithm for complex-valued data neural processing, which allows for separating mixed independent circular source signals. The proposed generalization relies on an interesting interpretation of nonclassical Hebbian learning proposed by Sudjianto and Hassoun (1994) for real-valued neural units.
Keywords :
Hebbian learning; neural nets; signal processing; blind separation; complex-valued data neural processing; complex-valued sources; extended Hebbian learning; generalization; mixed independent circular source signals; nonlinear extension; real-valued neural units; Computer simulation; Data mining; Hebbian theory; Neural networks; Neurons; Principal component analysis; Signal analysis; Signal processing; Signal processing algorithms; Source separation;
Journal_Title :
Signal Processing Letters, IEEE
