A New ICA Algorithm Based on Normalized Kurtosis

Cui Zhi-tao; Wen Jie-chang; Hu Nan-hui

English Version Journal Login

您当前的位置：

首页 >

文章列表页 >

A New ICA Algorithm Based on Normalized Kurtosis

更新时间：2026-05-27

- A New ICA Algorithm Based on Normalized Kurtosis
- Vol. 24, Issue 4, Pages: 38-41(2007)
- 作者机构：
- 作者简介：
- 基金信息：
- DOI：
  CLC：
- Online First：08 November 2007，
  
  Published：2007
- 稿件说明：
移动端阅览
崔志涛, 温洁嫦, 胡南辉. 一种新的基于归一化峭度的ICA算法[J]. 广东工业大学学报, 2007, 24(4): 38-41.

Cui Zhi-tao, Wen Jie-chang, Hu Nan-hui. A New ICA Algorithm Based on Normalized Kurtosis[J]. 2007, 24(4): 38-41.
崔志涛, 温洁嫦, 胡南辉. 一种新的基于归一化峭度的ICA算法[J]. 广东工业大学学报, 2007, 24(4): 38-41. DOI：

Cui Zhi-tao, Wen Jie-chang, Hu Nan-hui. A New ICA Algorithm Based on Normalized Kurtosis[J]. 2007, 24(4): 38-41. DOI：

摘要

基于峭度的独立分量分析中

采用梯度法寻找代价函数的极值点时

搜索的方向对算法的收敛起着关键的作用

对Kuicnet算法中原来的梯度做一个倍数替换

得到一个新的ICA算法.新算法在迭代过程中适时对梯度方向作出调整

加快收敛速度.数值仿真说明了算法的有效性. 更多还原

Abstract

The traditional approach to gradient to search for the cost function’s extreme in the independent component analysis was based on kurtosis.The path we are looking for is crucial to the convergence of the algorithm.In the Kuicnet algorithm

after replacing the gradient of the cost function with its scaling we obtained a new ICA algorithm.Adjusting the path of the gradient timely in the iterative process

it accelerates the convergence.The simulation illustrates the effectiveness.

关键词

Keywords

references

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

An Improved M ethod of Palmprint Recognition Based on Independent Component Analysis

The Cross-twin Physics-informed Neural Network for Wave Equations with Conserved Quantities

An Optimal Operation Strategy for Community Microgrid Alliances Considering Power Exchange under Multi-package Shared Energy Storage Model

Research on Foreign Object Detection in Power Transmission Lines Based on the AAGV-YOLOX Model

Bimodal Iterative Cross-attention Fusion Ensemble Framework

Related Author

PEI Yu

WEN Jie-Chang

Related Institution

No data

AI问答

⁰