Título: Non and semi-parametric estimation in models with unknown smoothness
Autores: Zinde-Walsh, Victoria
Kotlyarova, Yulia
Fecha: 2006
2006-06-07
Publicador: McGill University - MCGILL
Fuente:
Tipo: Text
Tema: Nonparametric estimation
Combined estimator
Descripción: Many asymptotic results for kernel-based estimators were established under some smoothness assumption on density. For cases where smoothness assumptions that are used to derive unbiasedness or asymptotic rate may not hold we propose a combined estimator that could lead to the best available rate without knowledge of density smoothness. A Monte Carlo example confirms good performance of the combined estimator.
Idioma: eng