The interpretation of Mallows's C-p-statistic

被引：78

作者：

Gilmour, SG ^{[1
]}

机构：

[1] UNIV READING, DEPT APPL STAT, READING RG6 6FN, BERKS, ENGLAND

来源：

JOURNAL OF THE ROYAL STATISTICAL SOCIETY SERIES D-THE STATISTICIAN | 1996年 / 45卷 / 01期

关键词：

multiple regression; multivariate F-distribution; variable selection;

D O I：

10.2307/2348411

中图分类号：

O21 [概率论与数理统计]; C8 [统计学];

学科分类号：

020208 ; 070103 ; 0714 ;

摘要：

When selecting variables in multiple-regression studies, the model with the lowest value of Mallows's C-p-statistic is often chosen. It is shown here that when the estimate of sigma(2) comes from the full model an adjusted C-p, (C) over bar(p), has the property that E((C) over bar(p)) = p. It Is suggested that a procedure be adopted which involves testing whether) It is model with minimum (C) over bar(p) is really better than a simpler model. Tables approximating the null distribution of the test statistics are given.

引用

页码：49 / 56

页数：8