Some results for robust GM-based estimators in heteroscedastic regression models

In this paper the asymptotic behavior of robust estimates based on GM-estimators when the observations follow a regression model with random carriers and heteroscedastic errors is established. Also one-step Newton-Raphson and reweighted estimates for these models are introduced and their breakdown p...

Descripción completa

Guardado en:
Detalles Bibliográficos
Autores principales: Bianco, A., Boente, G., Di Rienzo, J.
Formato: JOUR
Materias:
Asymptotic properties
Heteroscedasticity
High breakdown point estimates
Primary 62F35
Robust estimation
Secondary 62J05
Acceso en línea:http://hdl.handle.net/20.500.12110/paper_03783758_v89_n1-2_p215_Bianco
Aporte de:
Biblioteca Digital - Facultad de Ciencias Exactas y Naturales (UBA) de Universidad de Buenos Aires
Descripción
Sumario:In this paper the asymptotic behavior of robust estimates based on GM-estimators when the observations follow a regression model with random carriers and heteroscedastic errors is established. Also one-step Newton-Raphson and reweighted estimates for these models are introduced and their breakdown point is studied. Through a Monte Carlo study their performance for small samples is studied. © 2000 Elsevier Science B.V.

Ejemplares similares