| Paper: | MLSP-L4.5 | ||
| Session: | Blind Source Separation I | ||
| Time: | Thursday, May 18, 17:50 - 18:10 | ||
| Presentation: | Lecture | ||
| Topic: | Machine Learning for Signal Processing: Blind Signal Separation and Independent Component Analysis | ||
| Title: | Stability Analysis of Complex-valued Nonlinearities for Maximization of Nongaussianity | ||
| Authors: | Mike Novey, Tulay Adali, University of Maryland, Baltimore County, United States | ||
| Abstract: | Complex maximization of nongaussianity (CMN) has been shown to provide reliable separation of both circular and noncircular sources. It is also shown that the algorithm converges to the principal component of the source distribution when studied in the estimation direction. In this paper, we study the local stability of the CMN algorithm and determine the conditions under which local stability is achieved by extending our previous work to all dimensions of the weight vector. We use these conditions of stability to quantify convergence performance for a number of complex nonlinear functions, and present simulation results to demonstrate the effectiveness of these functions. | ||
©2018 Conference Management Services, Inc. -||- email: webmaster@icassp2006.org -||- Last updated Friday, August 17, 2012