Model averaging based on leave-subject-out cross-validation

This paper develops a frequentist model averaging method based on the leave-subject-out cross-validation. This method is applicable not only to averaging longitudinal data models, but also to averaging time series models which can have heteroscedastic errors. The resulting model averaging estimators...

Full description

Saved in:
Bibliographic Details
Published inJournal of econometrics Vol. 192; no. 1; pp. 139 - 151
Main Authors Gao, Yan, Zhang, Xinyu, Wang, Shouyang, Zou, Guohua
Format Journal Article
LanguageEnglish
Published Amsterdam Elsevier B.V 01.05.2016
Elsevier Sequoia S.A
Subjects
Asymptotic methods
Asymptotic optimality
econometric models
economic analysis
economic theory
Estimating techniques
heteroskedasticity
Leave-subject-out cross-validation
Longitudinal data
Mathematical models
Measurement errors
Model averaging
Simulation
Time series
time series analysis
Validation studies
C52
Model averaging
Time series
Asymptotic optimality
Leave-subject-out cross-validation
Longitudinal data
C51
Online AccessGet full text

Cover

More Information
Summary:This paper develops a frequentist model averaging method based on the leave-subject-out cross-validation. This method is applicable not only to averaging longitudinal data models, but also to averaging time series models which can have heteroscedastic errors. The resulting model averaging estimators are proved to be asymptotically optimal in the sense of achieving the lowest possible squared errors. Both simulation study and empirical example show the superiority of the proposed estimators over their competitors.
Bibliography:SourceType-Scholarly Journals-1
ObjectType-Feature-1
content type line 14
ObjectType-Article-1
ObjectType-Feature-2
content type line 23
ISSN:0304-4076
1872-6895
DOI:10.1016/j.jeconom.2015.07.006