摘要
基于另一种观点,本文对线性回归模型和自回归时间序列模型分别推导出AICC和AICC*选阶准则,同时对AICC准则在大样本情况选阶缺乏准确率的原因给予理论解释.
Based on a new point of view, criteria AICC and AICC are derived for linear model selection and autoregressive time series model selection respectively, and the reason of the unsatisfactory large sample performance of AICC is found. Simulations are made to show the different accuracies of the proposed criteria.
出处
《南开大学学报(自然科学版)》
CAS
CSCD
北大核心
1998年第4期43-49,111,共8页
Acta Scientiarum Naturalium Universitatis Nankaiensis