The Preprocessing Technique Of Independent Component Analysis: On The Approach Of Source Signals Ar Modeling

Shuxin Zheng,Weiqiang Zhang

2010-01-01

Abstract:Independent component analysis (ICA) is a statistical and computational technique for revealing hidden factors with statistical independence and to meet the non-Gaussian. Before applying an ICA algorithm, it is usually very useful to do some preprocessing, e.g., whitening is performed before applying the source separation algorithm because of that sources are often not white. However, the preprocessing techniques just like whitening, centering, factor analysis([1][2])and so on, that make the problem of ICA estimation simpler and better conditioned([3]).Since the AR models describe fairly well many natural stochastic processes, e.g., speech signals, they are used in many applications. This paper aims at. the preprocessing technique of ICA and especially focus on the performance of ICA based on the source signals AR modeling. It also presents the structure model of ICA based on the source signals preprocessing. Several simulation experimental results demonstrate that the source signals AR modeling can improve the performance of ICA algorithm on numerical stability.