Premium
PENALIZED‐ R 2 CRITERIA FOR MODEL SELECTION *
Author(s) -
TAYLOR LARRY W.
Publication year - 2009
Publication title -
the manchester school
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.361
H-Index - 42
eISSN - 1467-9957
pISSN - 1463-6786
DOI - 10.1111/j.1467-9957.2009.02123.x
Subject(s) - selection (genetic algorithm) , model selection , parametric statistics , econometrics , function (biology) , mathematical optimization , mathematics , least squares function approximation , parametric model , estimation , instrumental variable , dual (grammatical number) , sample (material) , computer science , mathematical economics , economics , statistics , artificial intelligence , art , chemistry , management , literature , chromatography , evolutionary biology , estimator , biology
It is beneficial to observe that popular model selection criteria for the linear model are equivalent to penalized versions of R 2 . Let PR 2 refer to any one of these model selection criteria. Then PR 2 serves the dual purpose of selecting the model and summarizing the resulting fit subject to the penalty function. Furthermore, it is straightforward to extend the logic of PR 2 to instrumental variables estimation and the non‐parametric selection of regressors. For two‐stage least squares estimation, a simulation study investigates the finite‐sample performance of PR 2 to select the correct model in cases of either strong or weak instruments.