MODEL BUILDING AND DATA MINING |
| |
Authors: | J. Denis Sargan |
| |
Affiliation: | a Department of Economics, London School of Economics, London, England |
| |
Abstract: | This paper defines the phenomenon of data mining in econometrics and discusses various outcomes of and solutions to data mining. Both classical and Bayesian approaches are considered, each with notable advantages and disadvantages, and with the choice of loss function affecting critical values. Illustrative examples include variable addition and exclusion in a standard linear regression model, the choice of lag structure in a dynamic single equation, and specification in a simultaneous equations model. |
| |
Keywords: | Bayes Loss function Pre-test estimation Specification searches Stein-James estimator JEL Classification: C44 C51 |
本文献已被 InformaWorld 等数据库收录! |
|