门限自回归模型参数估计的渐近性质

ASYMPTOTIC PROPERTIES OF ESTIMATES OF PARAMETERS FOR TAR MODELS

  • 摘要: 本文在给定门限自回归模型阶数、门限及延迟参数的情况下,通过研究模型所构成Markov链的一些性质,得到了自回归系数最小二乘法估计的强相容性与渐近正态性。同时,还证明了白噪声方差的估计亦具有强相容性。

     

    Abstract: In this paper, we study Marker properties of the Threshold Autoressive (TAR) Model in which the order, thresholds and delay parameter are given. Using these properties, we obtain the strong oonsistenoy and asymptotio normality of the least square estimates of the parameters for this model. The strong oonsistency of the estimates of varianoes of the white noises is presented, too.

     

/

返回文章
返回