时间序列建模中滞后阶数选取准则函数的检测效力及其特征

The Researches on the Test Power and Features on the Lagging Number Selecting Criteria about the Time Series Models

研究了时间序列建模中滞后阶数选取准则函数的检测效力及其特征.采用Monte-Carlo模拟对于所有的1阶和2阶自回归模型进行了系统仿真,研究结论表明:1)随着样本容量的逐渐增大,SIC准则函数正确判断出平稳序列滞后阶数的概率将显著增大,并渐近依概率收敛到1,而AIC准则函数正确判断出平稳序列滞后阶数的概率将逐渐增大,但并不能依概率收敛到真值.2)在其它外界条件不变情况下,所有准则函数正确判断概率与序列的平稳性无显著变化关系.3)样本容量越小,AICC检测效力越显著高于SIC、AIC检测效力;随着样本容量的逐渐增大,SIC检测效力逐渐高于AICC、AIC检测效力,但此时AICC、AIC检测效力相当.

The paper studies the test power and features on the lagging number selecting criteria about the time series models. The author uses Monte-Carlo methods to systematically simulation on the all lag 1 and lag 2 autoregressive models and the conclusions show： 1 ）The probability of correct judging of SIC increases obviously accompanying with the samples increasing and converges to 1 little by liUtle, but that of AIC can not converges to the true values. 2）The probabilities of correct judging by all the criteria functions have no relations to the stabilities of the variables. 3） The test power of AICC is much higher than that of AIC and SIC when the sample is small, which of SIC is higher than that of AIC and AICC little by little as the samples increase, furthermore the test power of AIC nearUy equals to that of AICC finally.