ScholarLens Back to search
2005•Journal of North China Institute of TechnologyRequires access

An Independent Component Analysis Algorithm Based on Maximum Negentropy

Shen Li-yan

Open publisher page 0 citations

Abstract

The independent component analysis(ICA) is a method widely used in blinding source separation and is developed in recent years.The basic principle of the ICA is discussed in this paper.Using maximum negentropy approximations of differential entropy,we introduce an objective function for ICA and present a FastICA algorithm based on maximum negentropy.Simulation experiments in MATLAB show its better performance and efficiency in mixed sound signal separation,proving its good convergence and robust.

About this research paper

What this paper is about

The independent component analysis(ICA) is a method widely used in blinding source separation and is developed in recent years.The basic principle of the ICA is discussed in this paper.Using maximum negentropy approximations of differential entropy,we introduce an objective function for ICA and present a FastICA algorithm based on maximum negentropy.Simulation experiments in MATLAB show its better performance and efficiency in mixed sound signal separation,proving its good convergence and robust.

Why it matters

A significance statement is not available in the OpenAlex record.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

The independent component analysis(ICA) is a method widely used in blinding source separation and is developed in recent years.The basic principle of the ICA is discussed in this paper.Using maximum negentropy approximations of differential entropy,we introduce an objective function for ICA and present a FastICA algorithm based on maximum negentropy.Simulation experiments in MATLAB show its better performance and efficiency in mixed sound signal separation,proving its good convergence and robust.

Key concepts: Negentropy, Independent component analysis, FastICA, Blind signal separation, Algorithm, Entropy (arrow of time), Computer science, MATLAB

Related papers

Related paper 1Related paper 2Related paper 3
Back to paper searchBrowse research topicsOriginal source
An Independent Component Analysis Algorithm Based on Maximum Negentropy — Research Paper | ScholarLens