多维门限自回归模型参数估计的强相容性  

Strong Consistency of Estimates of Parameters for Multiple TAR Models

作　　者：宋心远[1] 邓集贤[1] 

机构地区：[1]中山大学数学系

出　　处：《中山大学学报（自然科学版）》1990年第1期23-28,共6页Acta Scientiarum Naturalium Universitatis Sunyatseni

基　　金：国家教委博士点基金

摘　　要：对多维门限自回归模型在给定阶数、门限及延迟参数的假定下,通过研究模型所构成的Markov链的遍历性,得到了自回归系数及白噪声协方差阵最小二乘方估计的强相容性。Through studying markov properties of the multiple Threshold Autoregressi-re(TAR)Model in which the order,the thresholds and delay parameter are giv-en,we obtain the strong consistency of the least square estimates of the coeffic-int matrices and covariance matrix of the white noises.The results extend the corresponding results of paper(6).

关 键 词：门限自回归 相容性 遍历性 

分 类 号：O212.1[理学—概率论与数理统计]