Premium
Model Selection in Equations with Many ‘Small’ Effects *
Author(s) -
Castle Jennifer L.,
Doornik Jurgen A.,
Hendry David F.
Publication year - 2013
Publication title -
oxford bulletin of economics and statistics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.131
H-Index - 73
eISSN - 1468-0084
pISSN - 0305-9049
DOI - 10.1111/j.1468-0084.2012.00727.x
Subject(s) - collinearity , selection (genetic algorithm) , salient , dimension (graph theory) , econometrics , model selection , computer science , dimensionality reduction , nonlinear system , principal component analysis , feature selection , principal (computer security) , mathematics , statistics , machine learning , artificial intelligence , physics , quantum mechanics , pure mathematics , operating system
High dimensional general unrestricted models (GUMs) may include important individual determinants, many small relevant effects, and irrelevant variables. Automatic model selection procedures can handle more candidate variables than observations, allowing substantial dimension reduction from GUMs with salient regressors, lags, nonlinear transformations, and multiple location shifts, together with all the principal components, possibly representing ‘factor’ structures, as perfect collinearity is also unproblematic. ‘Factors’ can capture small influences that selection may not retain individually. The final model can implicitly include more variables than observations, entering via ‘factors’. We simulate selection in several special cases to illustrate.