ICA by mutual information minimization: An approach for avoiding local minima

Using Mutual Information (MI) minimization is very common in Blind Source Separation (BSS). However, it is known that gradient descent approaches may trap in local minima of MI in constrained models. In this paper, it is proposed that this problem may be solved using a `poor' estimation of the derivative of MI.

[1]  Christian Jutten,et al.  A general approach for mutual information minimization and its application to blind source separation , 2005, Signal Process..

[2]  Dinh-Tuan Pham,et al.  Mutual information approach to blind separation of stationary sources , 2002, IEEE Trans. Inf. Theory.

[3]  Christian Jutten,et al.  Differential of the mutual information , 2004, IEEE Signal Processing Letters.

[4]  D. Cox.  A penalty method for nonparametric estimation of the logarithmic derivative of a density function , 1985 .

[5]  Christian Jutten,et al.  Space or time adaptive signal processing by neural network models , 1987 .

[6]  John G. Proakis,et al.  Probability, random variables and stochastic processes , 1985, IEEE Trans. Acoust. Speech Signal Process..

[7]  E. Oja,et al.  Independent Component Analysis , 2013 .

[8]  Andrzej Cichocki,et al.  Adaptive Blind Signal and Image Processing - Learning Algorithms and Applications , 2002 .

[9]  Jean-Francois Cardoso,et al.  Blind signal separation: statistical principles , 1998, Proc. IEEE.

[10]  Pierre Comon,et al.  Independent component analysis, A new concept? , 1994, Signal Process..

[11]  P. Comon.  Independent Component Analysis , 1992 .

[12]  Christian Jutten,et al.  Entropy Optimization - Application to Blind Source Separation , 1997, ICANN.

[13]  Michel Verleysen,et al.  On the entropy minimization of a linear mixture of variables for source separation , 2005, Signal Process..

[14]  C. D. Kemp,et al.  Density Estimation for Statistics and Data Analysis , 1987 .