Convergence Analysis for a Class of Source Separation Methods
 |
Prof. Alper T. Erdogan,Koc University, Istanbul, Turkey |
Description: Blind source separation(BSS) is a topic of interest in different fields of information processing with a large span of applications, e.g., in communications, pattern recognition, fMRI and audio processing. This talk will introduce a new source separation scheme called Bounded Component Analysis (BCA) where the boundedness of sources are exploited to replace the independence assumption, used by Independent Component Analysis (ICA), with a weaker statistical assumption. The focus will be on the presentation of an algorithmic framework for BCA as well as relevant convergence results.
Organizers
Techical co-sponosors
Supporters:
General sponsor
Principal sponsor
Supporters:
