Performance comparison of new nonparametric independent component analysis algorithm for different entropic indexes

Authors : Yusuf SEVİM, Ayten ATASOY

Pages : 287-297

View : 12 | Download : 7

Publication Date : 0000-00-00

Article Type : Research Paper

Abstract :Most independent component analysis insert ignore into journalissuearticles values(ICA); algorithms use mutual information insert ignore into journalissuearticles values(MI); measures based on Shannon entropy as a cost function, but Shannon entropy is not the only measure in the literature. In this paper, instead of Shannon entropy, Tsallis entropy is used and a novel ICA algorithm, which uses kernel density estimation insert ignore into journalissuearticles values(KDE); for estimation of source distributions, is proposed. KDE is directly evaluated from the original data samples, so it solves the important problem in ICA: how to choose nonlinear functions as the probability density function insert ignore into journalissuearticles values(pdf); estimation of the sources.
Keywords : Independent component analysis ICA, kernel density estimation KDE, shannon entropy, tsallis entropy, mutualinformation insert ignore into journalissuearticle