|
|
|
| Improved Adaptive Algorithm of Blind Source Separation Based on Nonholonomic Natural Gradient |
| NIU YiLong1, WANG YingMin1, WANG Yi2 |
1.College of Marine, Northwestern Polytechnic University, Xi’an 710072
2.School of Electronic and Information, Northwestern Polytechnic University, Xi’an 710072 |
|
|
|
|
Abstract Compared with the natural gradient learning algorithm of blind source separation, the nonholonomic natural gradient algorithm avoids numerical instability which is caused by the nonstationary source signals and the rapid magnitude change. Aiming at the difficulty in determining the nonlinear activation function, an improved algorithm using kurtosis to select activation function adaptively without available prior information is proposed. It retains the predominance of nonholonomic natural gradient algorithm in restoring nonstationary sources, and can be adapted to the sources of arbitrary distribution. Computer simulations show the performance of proposed method is better than that of the original algorithm with tangent function.
|
|
Received: 09 June 2005
|
|
|
|
|
|
[1] Cichocki A, Unbehauen R, Moszczyński L, et al. A New On-Line Adaptive Learning Algorithm for Blind Separation of Sources // Proc of the International Joint Conference on Neural Networks. Taiwan, China, 1994: 406-411
[2] Amari S. Natural Gradient Works Efficiently in Learning. Neural Computation, 1998, 10(2): 251-276
[3] Zhang L Q, Amari S, Cichocki A. Natural Gradient Approach to Blind Separation of Over- and Under-Complete Mixtures // Proc of the International Workshop on Independent Component Analysis and Blind Signal Separation. Aussois, France, 1999: 455-460
[4] Amari S. Natural Gradient Learning for Over- and Under-Complete Bases in ICA. Neural Computation, 1999, 11(8): 1875-1883
[5] Amari S, Chen T P, Cichocki A. Non-Holonomic Constraints in Learning Algorithms for Blind Source Separation. Neural Computation, 2000, 12(7): 1463-1484
[6] Bell A J, Sejnowski T J. An Information Maximization Approach to Blind Separation and Blind Deconvolution. Neural Computation, 1995, 7(6): 1129-1159
[7] Yang H H, Amari S. A Stochastic Natural Gradient Descent Algorithm for Blind Signal Separation // Proc of the IEEE Workshop on Neural Networks for Signal Processing. Kyoto, Japan, 1997: 436-444
[8] Comon P. Independent Component Analysis, A New Concept? Signal Processing, 1994, 36(3): 287-314
[9] Cichocki A, Amari S. Adaptive Blind Signal and Image Processing: Learning Algorithm and Applications. Chichester, UK: John Wiley, 2002 |
|
|
|
2006, Vol. 19 
