| 基于随机数据取样技术的FASTICA算法 |
| |
| 引用本文： | 李艳丽,杜锋. 基于随机数据取样技术的FASTICA算法[J]. 湖北大学学报(自然科学版), 2011, 33(4): 490-493 |
| |
| 作者姓名： | 李艳丽 杜锋 |
| |
| 作者单位： | 1. 荆楚理工学院电子信息工程学院,湖北荆门,448000
2. 湖北大学数学与计算机科学学院,湖北武汉430062;荆楚理工学院数理学院,湖北荆门448000 |
| |
| 基金项目： | 荆楚理工学院科学研究项目，应用数学湖北省重点实验室资助 |
| |
| 摘 要： | 提出一种随机数据取样的方法,通过在大量的原始数据中随机选取一部分进行分析,在不影响分离效果的前提下,使得FASTICA所需要的时间大为减少.利用峭度估计器分析在一定的置信区间和置信水平的条件下得到取样比例的下限.计算机仿真结果证明这种取样技术的有效性,并且分析不同取样比例下的FASTICA算法性能.
|
| 关 键 词： | 数据取样 峭度 FASTICA |
A FASTICA algorithm based on random data sampling |
| |
| LI Yanli , DU Feng. A FASTICA algorithm based on random data sampling[J]. Journal of Hubei University(Natural Science Edition), 2011, 33(4): 490-493 |
| |
| Authors: | LI Yanli DU Feng |
| |
| Affiliation: | 1.Department of Electronic and Information,Jingchu University of Technology,Jingmen 448000,China; 2.School of Mathematics and Computer Science,Hubei University,Wuhan 430062,China; 3.School of Mathematics and Physics,Jingchu University of Technology,Jingmen 448000,China) |
| |
| Abstract: | A method of random sampling of data was given,which was based on analysis of random selecting partly data from a large amount of raw data.Without prejudice to the effect of blind source separation,the method greatly reduced the time that FASTICA algorithm required.The ratio threshold was founded by performing an analysis of the kurtosis estimator with a fixed confidence interval and confidence level.Computer simulation results showed that the validation of this method,and analyzed algorithm performances with different reduction ratio. |
| |
| Keywords: | data sampling kurtosis FASTICA |
| 本文献已被 维普 万方数据 等数据库收录！ |
|
