Variable deletion conficence regions and bootstrapping in linear regression

Velilla Cerdan, Santiago

e-Archivo Principal
→
Investigación
→
Departamentos
→
Departamento de Estadística
→
DES - Working Papers. Statistics and Econometrics. WS
→
Ver ítem

Variable deletion conficence regions and bootstrapping in linear regression

Autor(es): Velilla Cerdan, Santiago

Editor: Universidad Carlos III de Madrid. Departamento de Estadística

Fecha de edición: 1999-10

Serie/Num.: UC3M Working Papers. Statistics and Econometrics
99-72-25

Palabras clave: Least squeres estimation , mallows distance , model selection , resampling

URI: http://hdl.handle.net/10016/6351

Derechos: Atribución-NoComercial-SinDerivadas 3.0 España

Resumen:

A resampling method is introduced to approximate, when some of the predictors are deleted, the quantiles of the distribution of the usual least squares pivots in linear regression. The approximation is used to construct confidence regions for the parameters of [+]

Mostrar el registro completo del ítem