Performance analysis of the FastICA algorithm and Crame/spl acute/r-rao bounds for linear independent component analysis
Authors:
- Z. KoldovskyZKZ. Koldovsky
- E. OjaEOE. Oja
- 20 March 2006
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Signal Processing
- Vol. 54 (4) , 1189-1203
Abstract
The FastICA or fixed-point algorithm is one of the most successful algorithms for linear independent component analysis (ICA) in terms of accuracy and computational complexity. Two versions of the algorithm are available in literature and software: a one-unit (deflation) algorithm and a symmetric algorithm. The main result of this paper are analytic closed-form expressions that characterize the separating ability of both versions of the algorithm in a local sense, assuming a "good" initialization of the algorithms and long data records. Based on the analysis, it is possible to combine the advantages of the symmetric and one-unit version algorithms and predict their performance. To validate the analysis, a simple check of saddle points of the cost function is proposed that allows to find a global minimum of the cost function in almost 100% simulation runs. Second, the Crame/spl acute/r-Rao lower bound for linear ICA is derived as an algorithm independent limit of the achievable separation quality. The FastICA algorithm is shown to approach this limit in certain scenarios. Extensive computer simulations supporting the theoretical findings are included.Keywords
This publication has 32 references indexed in Scilit:
- QML Blind Deconvolution: Asymptotic Analysis2004
- A blind-ML scheme for blind source separation2004
- Convergence of the symmetrical FastICA algorithm2004
- Monotonic convergence of fixed-point algorithms for ICAIEEE Transactions on Neural Networks, 2003
- One-unit contrast functions for independent component analysis: a statistical analysis2002
- Globally convergent blind source separation based on a multiuser kurtosis maximization criterionIEEE Transactions on Signal Processing, 2000
- A Fast Fixed-Point Algorithm for Independent Component Analysis of Complex Valued SignalsInternational Journal of Neural Systems, 2000
- AN EXPERIMENTAL COMPARISON OF NEURAL ALGORITHMS FOR INDEPENDENT COMPONENT ANALYSIS AND BLIND SEPARATIONInternational Journal of Neural Systems, 1999
- Equivariant adaptive source separationIEEE Transactions on Signal Processing, 1996
- Linear Statistical Inference and its Applications1973