基于八阶收敛牛顿迭代的Fast-ICA改进算法

发布时间：2018-03-08 13:51

本文选题：盲源分离(BSS)　切入点：独立分量分析(ICA)　出处：《计算机工程与应用》2017年11期 　论文类型：期刊论文

【摘要】：解决盲源分离问题(BSS)最常用的方法是独立分量分析方法(ICA),快速独立分量分析方法(Fast-ICA)是目前广泛使用的独立分量分析方法。传统的Fast-ICA算法利用了二阶收敛的牛顿迭代方法进行优化,为了加快算法的收敛速度,提高算法的运行效率,利用八阶收敛的牛顿迭代方法对Fast-ICA算法进行优化,通过仿真验证了基于八阶收敛的Fast-ICA算法与传统的Fast-ICA和五阶收敛的Fast-ICA算法在分离性能上基本相同,但其具有更少的迭代次数和更快的收敛速率。
[Abstract]:The most commonly used method to solve blind source separation problem is the independent component analysis (ICA), the fast independent component analysis (Fast-ICA), which is widely used at present. The traditional Fast-ICA algorithm uses the second-order convergent Newtonian method. To optimize the method, In order to speed up the convergence of the algorithm and improve the efficiency of the algorithm, the eighth order convergent Newton iteration method is used to optimize the Fast-ICA algorithm. The simulation results show that the Fast-ICA algorithm based on eighth-order convergence has the same separation performance as the traditional Fast-ICA and five-order convergent Fast-ICA algorithm, but it has fewer iterations and faster convergence rate.
【作者单位】：湖南大学信息科学与工程学院;
【基金】：湖南省高校创新平台开放基金项目(No.14K022)
【分类号】：TN911.7

【相似文献】