Correcting the corrected AIC

被引：10

作者：

DelSole, Timothy ^{[1
]}

Tippett, Michael K. ^{[2
]}

机构：

[1] George Mason Univ, Dept Atmospher Ocean & Earth Sci, Fairfax, VA 22030 USA

[2] Columbia Univ, Dept Appl Phys & Appl Math, New York, NY USA

来源：

STATISTICS & PROBABILITY LETTERS | 2021年 / 173卷

基金：

美国国家科学基金会; 美国国家航空航天局; 美国海洋和大气管理局;

关键词：

Model selection; Akaike's Information Criterion; AICc; ANCOVA; MODEL SELECTION; REGRESSION;

D O I：

10.1016/j.spl.2021.109064

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：

The standard correction to Akaike's Information Criterion, AICc, assumes the same predictors for training and verification and therefore underestimates prediction error for random predictors. A corrected AIC for regression models containing a mix of random and fixed predictors is derived. (C) 2021 The Author(s). Published by Elsevier B.V.

引用

页数：7

共 13 条

[1] NEW LOOK AT STATISTICAL-MODEL IDENTIFICATION [J].

AKAIKE, H .

IEEE TRANSACTIONS ON AUTOMATIC CONTROL, 1974, AC19 (06) :716-723

[2]

[Anonymous], 1985, MULT AN 6 P 6 INT S

[3]

[Anonymous], 2009, ASPECTS MULTIVARIATE

[4] MODEL SELECTION FOR MULTIVARIATE REGRESSION IN SMALL SAMPLES [J].

BEDRICK, EJ ;

TSAI, CL .

BIOMETRICS, 1994, 50 (01) :226-231

[5]

Brockwell P.J., 2006, Introduction to time series and forecasting, V3rd

[6]

Burnham K. P., 2003, MODEL SELECTION MULT, DOI DOI 10.1007/B97636

[7]

DelSole T., 2020, STAT

[8] REGRESSION AND TIME-SERIES MODEL SELECTION IN SMALL SAMPLES [J].

HURVICH, CM ;

TSAI, CL .

BIOMETRIKA, 1989, 76 (02) :297-307

[9]

Mardia K. V., 1979, Multivariate Analysis. Probability and Mathematical Statistics, P457

[10]

McQuarrie A. D. R., 1998, Regression and Time Series Model Selection

← 1 2 →