PASCAL - Pattern Analysis, Statistical Modelling and Computational Learning

Performance analysis of the FAstICA algorithm and Cramer-Rao lower bounds for linear independent component analysis
Petr Tichavsky, Zbynek Koldovski and Erkki Oja
IEEE Transactions on Signal Processing Volume 54, Number 4, pp. 1189-1203, 2006.

Abstract

The Cramer-Rao lower bound is derived for linear independent component analysis, and the performance of the FastICA algorithm is compared to this optimal bound, with good results in some special cases.

EPrint Type:Article
Project Keyword:Project Keyword UNSPECIFIED
Subjects:Computational, Information-Theoretic Learning with Statistics
Theory & Algorithms
ID Code:2551
Deposited By:Erkki Oja
Deposited On:22 November 2006