摘要
利用CR微分理论,提出快速主成分分析(Fast ICA)算法一种完全基于复域上的新的推导方法,避免了原推导方法中因复值数据拆分引起的推导过程繁琐等不足,更易于算法的扩展.
By using CR calculus theory,a new derivation of Fast ICA algorithm is presented which can be carried out in the complex domain completely. Due to avoiding splitting complex data into real and imaginary parts,the new derivation has some advantage of less cumbersome in expressions than the original derivation and is more convenient for extension of Fast ICA algorithm.
出处
《闽江学院学报》
2016年第2期31-34,共4页
Journal of Minjiang University
基金
福建省中青年教师教育科研项目(JA15436)
关键词
FASTICA算法
CR微分
复变量优化问题
牛顿法
FastICA algorithm
CR calculus theory
complex variables optimization problem
Newton method