Stable spectral factorization with applications to the estimation of time series models
Stable spectral factorization with applications to the estimation of time series models
复制标题
稳定谱分解及其在时间序列模型估计中的应用
DOI：
10.1080/03610929308831024
复制
发表时间：
1993
影响因子：
0.8
通讯作者：
M. Salau
中科院分区：
文献类型：
--
作者：
D. Poskitt;M. Salau
提出了一种由不稳定的v×v满秩多项式矩阵产生微小相位和不可逆谱因子的方法。单位圆内的零点使用闭合形式的代数运算通过边界|z|=1反映出来。程序中还包括一种确定多项式算子稳定性的技术，它不需要显式构造行列式方程。文中说明了该技术的应用，并讨论了该方法在系统估计的统计背景下的实现。
A method for generating a miniphase and inveitible spectral factor from an unstable v × v full rank polynomial matrix is proposed. The zeros inside the unit circle are reflected through the boundary |z|=1 using closed form algebraic manipulations. Also included in the procedure is a technique foi determining the stability of a polynomial operator that does not require the explicit construction of the determinant al equation. Application of the technique is illustrated and the implementation of the method in the statistical context of system estimation is discussed.