On the Convergence of ICA Algorithms With Symmetric Orthogonalization
- 10 February 2009
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Signal Processing
- Vol. 57 (6), 2209-2221
- https://doi.org/10.1109/tsp.2009.2015114
Abstract
Independent component analysis (ICA) problem is often posed as the maximization/minimization of an objective/cost function under a unitary constraint, which presumes the prewhitening of the observed mixtures. The parallel adaptive algorithms corresponding to this optimization setting, where all the separators are jointly trained, are typically implemented by a gradient-based update of the separation matrix followed by the so-called symmetrical orthogonalization procedure to impose the unitary constraint. This article addresses the convergence analysis of such algorithms, which has been considered as a difficult task due to the complication caused by the minimum-(Frobenius or induced 2-norm) distance mapping step. We first provide a general characterization of the stationary points corresponding to these algorithms. Furthermore, we show that fixed point algorithms employing symmetrical orthogonalization are monotonically convergent for convex objective functions. We later generalize this convergence result for nonconvex objective functions. At the last part of the article, we concentrate on the kurtosis objective function as a special case. We provide a new set of critical points based on Householder reflection and we also provide the analysis for the minima/maxima/saddle-point classification of these critical points.Keywords
This publication has 22 references indexed in Scilit:
- Optimization Algorithms on Matrix ManifoldsPublished by Walter de Gruyter GmbH ,2008
- On the convergence of ICA algorithms with symmetric orthogonalizationPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2008
- Complex-Valued Matrix Differentiation: Techniques and Key ResultsIEEE Transactions on Signal Processing, 2007
- Unified Theory of Complex-Valued Matrix DifferentiationPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2007
- Gradient and Fixed-Point Complex ICA Algorithms Based on Kurtosis Maximization2007 IEEE Workshop on Machine Learning for Signal Processing, 2006
- Optimization in the Complex Domain for Nonlinear Adaptive FilteringPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2006
- Monotonic convergence of fixed-point algorithms for ICAIEEE Transactions on Neural Networks, 2003
- Optimization algorithms exploiting unitary constraintsIEEE Transactions on Signal Processing, 2002
- Globally convergent blind source separation based on a multiuser kurtosis maximization criterionIEEE Transactions on Signal Processing, 2000
- Complex gradient and HessianIEE Proceedings - Vision, Image, and Signal Processing, 1994