MALLOWS CP CRITERION AND UNBIASEDNESS OF MODEL SELECTION

被引：26

作者：

KOBAYASHI, M ^{[1
]}

SAKATA, S ^{[1
]}

机构：

[1] DOSHISHA UNIV,KYOTO 602,JAPAN

来源：

JOURNAL OF ECONOMETRICS | 1990年 / 45卷 / 03期

关键词：

D O I：

10.1016/0304-4076(90)90006-F

中图分类号：

F [经济];

学科分类号：

02 ;

摘要：

The model selection based on Mallows' Cp criterion is biased in the sense that the probability of selecting each from two linear models is not equal to 0.5 even if they have equal mean squared prediction errors. The authors propose a bias correction of the Mallows' Cp rule by means of the second-order asymptotic distribution of the difference of Mallows' Cp statistic. The authors also show that the proposed correction is necessary only if the difference of Mallows' Cp statistic is smaller than one in absolute value. © 1990.

引用

页码：385 / 395

页数：11

共 11 条

[1] FITTING AUTOREGRESSIVE MODELS FOR PREDICTION [J].