Constrained ICA and applications

The thesis begins with proposing a unified approach to perform complete ICA, the uICA, for Blind Source Separation (BSS). Three major ICA algorithms are investigated and shown to fall into this approach. They differ only from the nonlinear functions used in the learning rules.

Saved in:
Bibliographic Details
Main Author: Lu, Wei.
Other Authors: Rajapakse, Jagath Chandana
Format: Theses and Dissertations
Published: 2008
Subjects:
Online Access:http://hdl.handle.net/10356/2382
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Description
Summary:The thesis begins with proposing a unified approach to perform complete ICA, the uICA, for Blind Source Separation (BSS). Three major ICA algorithms are investigated and shown to fall into this approach. They differ only from the nonlinear functions used in the learning rules.