Independent component analysis by the PFANN neural network

The aim of this paper is to present a new neural technique for performing Independent Component Analysis of eterokurtic signals by the functional-link network. Based upon entropy optimization, the proposed approach relies on the use of a neural network formed by neural units endowed with adaptive activation functions that allow the recursive approximation of the cumulative distribution functions of the sources, that have been proved to be optimal in blind separation. Through computer simulations we show the proposed algorithm is effective and may exhibit performances similar to those of closely related algorithms found in the literature, requiring less computational effort.