Phân tách nguồn mù và kỹ thuật phân tách thành phần độc lập

Abstract

Blind source separation (BSS) problem and independent component analysic (ICA) techniques for performing BSS which seek to uncover the independent source signals from a set of observations that are composed of linear mixtures of the underlying sources. These techniques utilize a representation of the data in statistical domain rather than a time or frequency domain. In this document we will discuss interactive methods are used to maximize or minimize a given cost function such as mutual information, entropy or fourth order moment (kurtosis).