Abstract
We propose an independent component analysis (ICA) algorithm which can separate mixtures of sub- and super- Gaussian source signals with self-adaptive nonlinearities. The ICA algorithm in the framework of natural Riemannian gradient, is derived using the parameterized Generalized Compound Gamma Distribution density model. The nonlinear function in ICA algorithem is self-adaptive and is controlled by the shape parameter of Adaptive Generalized Compound Gamma Distribution density model. Computer simulation results confirm the validity and high performance of the proposed algorithm
Highlights
Maximum Entropy Algorithm3307| P a g e block diagram in Figure 1 explains the maximum entropy method for blind source separation
Given X, the problem is to find a demixing matrix W such that the original source vector S can be recovered from the output vector Y defined by Y=WX where Y [Y1,Y2,...,Ym ]T
The maximum entropy and maximum likelihood methods for blind source separation are equivalent under the condition that the random variable i z is uniformly distributed inside the interval [0,1] for all i
Summary
3307| P a g e block diagram in Figure 1 explains the maximum entropy method for blind source separation. The results from Maximum Entropy Method may be stated as follows: Let the non-linearity at the demixer output be defined in terms of the original source distribution as 3308| P a g e zi z i g i ( yi ) fS i (si )dsi , for all i = 1,2,...,n. The maximum entropy and maximum likelihood methods for blind source separation are equivalent under the condition that the random variable i z is uniformly distributed inside the interval [0,1] for all i. This relationship may be proven with the help of chain rule of calculus as n zi k 1 zi yi xi xi si n k 1. It is sensitive to the learning rate parameter and works better for super-Gaussian signals
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
More From: INTERNATIONAL JOURNAL OF COMPUTERS & TECHNOLOGY
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.