Model Selection in Equations with Many 'Small' Effects.
General unrestricted models (GUMs) may include important individual determinants, many small relevant effects, and irrelevant variables. Automatic model selection procedures can handle perfect collinearity and more candidate variables than observations, allowing substantial dimension reduction from...
| Main Authors: | , , |
|---|---|
| Format: | Working paper |
| Language: | English |
| Published: |
Department of Economics (University of Oxford)
2011
|
| _version_ | 1826268735919357952 |
|---|---|
| author | Castle, J Doornik, J Hendry, D |
| author_facet | Castle, J Doornik, J Hendry, D |
| author_sort | Castle, J |
| collection | OXFORD |
| description | General unrestricted models (GUMs) may include important individual determinants, many small relevant effects, and irrelevant variables. Automatic model selection procedures can handle perfect collinearity and more candidate variables than observations, allowing substantial dimension reduction from GUMs with salient regressors, lags, non-linear transformations, and multiple location shifts, together with all the principal components representing ‘factor’ structures, which can also capture small influences that selection may not retain individually. High dimensional GUMs and even the final model can implicitly include more variables than observations entering via ‘factors’. We simulate selection in several special cases to illustrate. |
| first_indexed | 2024-03-06T21:14:10Z |
| format | Working paper |
| id | oxford-uuid:3f340277-d1e8-45fd-80f4-89a94d857493 |
| institution | University of Oxford |
| language | English |
| last_indexed | 2024-03-06T21:14:10Z |
| publishDate | 2011 |
| publisher | Department of Economics (University of Oxford) |
| record_format | dspace |
| spelling | oxford-uuid:3f340277-d1e8-45fd-80f4-89a94d8574932022-03-26T14:30:28ZModel Selection in Equations with Many 'Small' Effects.Working paperhttp://purl.org/coar/resource_type/c_8042uuid:3f340277-d1e8-45fd-80f4-89a94d857493EnglishDepartment of Economics - ePrintsDepartment of Economics (University of Oxford)2011Castle, JDoornik, JHendry, DGeneral unrestricted models (GUMs) may include important individual determinants, many small relevant effects, and irrelevant variables. Automatic model selection procedures can handle perfect collinearity and more candidate variables than observations, allowing substantial dimension reduction from GUMs with salient regressors, lags, non-linear transformations, and multiple location shifts, together with all the principal components representing ‘factor’ structures, which can also capture small influences that selection may not retain individually. High dimensional GUMs and even the final model can implicitly include more variables than observations entering via ‘factors’. We simulate selection in several special cases to illustrate. |
| spellingShingle | Castle, J Doornik, J Hendry, D Model Selection in Equations with Many 'Small' Effects. |
| title | Model Selection in Equations with Many 'Small' Effects. |
| title_full | Model Selection in Equations with Many 'Small' Effects. |
| title_fullStr | Model Selection in Equations with Many 'Small' Effects. |
| title_full_unstemmed | Model Selection in Equations with Many 'Small' Effects. |
| title_short | Model Selection in Equations with Many 'Small' Effects. |
| title_sort | model selection in equations with many small effects |
| work_keys_str_mv | AT castlej modelselectioninequationswithmanysmalleffects AT doornikj modelselectioninequationswithmanysmalleffects AT hendryd modelselectioninequationswithmanysmalleffects |