摘要
基于矩阵的广义逆,本文给出了关于矩阵序列加速加速收敛的三个有理外推方法。它们包括:(i)基于广义逆的矩阵Pade逼近[4];(ii)矩阵Epsilon算法;(iii)矩阵AitkenΔ^2-算法。对三种方法之间的内在联系进行了讨论。关于Markov过程的一个实例给出以说明本文的结果。
Three rational extrapolation methods for accelerating the convergence of a sequence of matrix. based on using generalized inverse, are described. The underlying theory involves: (l)generalized inverse matrix valued Pade approximants; (ii)matrix valued Epsilon algorithm; (iii)matrix valued Ait.kenΔ2- method. Improtant inner connection among three methods is investigated. A useful example of Markov process is given so as to illustrate the results in this paper.
出处
《应用数学与计算数学学报》
1998年第2期46-52,共7页
Communication on Applied Mathematics and Computation
基金
国家自然科学基金资助项目