Modelling Nonlinearity and Long Memory in Time Series - (Now published in 'Nonlinear Dynamics and Time Series', C D Cutler and D T Kaplan (eds), Fields Institute Communications, 11 (1997), pp.61-170.)
Modelling Nonlinearity and Long Memory in Time Series - (Now published in 'Nonlinear Dynamics and Time Series', C D Cutler and D T Kaplan (eds), Fields Institute Communications, 11 (1997), pp.61-170.)
复制标题
时间序列中的非线性和长记忆建模 -（现已发表于“非线性动力学和时间序列”，C D Cutler 和 D T Kaplan（编辑），Fields Institute Communications，11 (1997)，第 61-170 页。）
DOI：
10.1090/fic/011/11
复制
发表时间：
1997
期刊：
影响因子：
--
通讯作者：
P. Zaffaroni
中科院分区：
文献类型：
--
作者：
P. Robinson;P. Zaffaroni
我们讨论在原始时间序列 xt 或其瞬时函数中赋予某种形式的长记忆的模型，特别是 。基于线性或非线性模型。讨论了 xt 的线性模型在 xt 的非线性函数中暗示长记忆的能力。经验观察激发了对模型的研究，这些模型导致短记忆，甚至白噪声，但长记忆。我们描述的这样一个模型是基于 Robinson (1991b) 引入的长记忆广义 ARCH 模型。另一个是Robinson (1977)非线性移动平均模型的扩展。
We discuss models that impart a form of long memory in raw time series xt or instantaneous functions thereof, in particular . on the basis of a linear or nonlinear model. The capacity of linear models for xt to imply long-memory in nonlinear functions of xt is discussed. Empirical observation motivates investigation of models which lead to short memory, or even white noise, xt but a long memory . One such model which we describe is based on the long memory generalized ARCH model introduced by Robinson (1991b). The other is an extension of the nonlinear moving average model of Robinson (1977).