Comparison of presmoothing methods in kernel density estimation under censoring

Authors:
M. A. Jácome;I. Gijbels;R. Cao
Affiliations:
Departamento de Matemáticas, Facultad de Ciencias, Universidade da Coruña, A Coruña, Spain 15071;Department of Mathematics and University Center for Statistics, Katholieke Universiteit Leuven, Leuven (Heverlee), Belgium 3001;Departamento de Matemáticas, Facultad de Informática, Universidade da Coruña, A Coruña, Spain 15071
Venue:
Computational Statistics
Year:
2008

Citing 0
Cited 2

A kernel-based parametric method for conditional density estimation

Pattern Recognition
Bayesian adaptive bandwidth kernel density estimation of irregular multivariate distributions

Computational Statistics & Data Analysis

Quantified Score

Hi-index	0.00

Visualization

Abstract

The behavior of the presmoothed density estimator is studied when different ways to estimate the conditional probability of uncensoring are used. The Nadaraya---Watson, local linear and local logistic approach are compared via simulations with the classical Kaplan---Meier estimator. While the local logistic presmoothing estimator presents the best performance, the relative benefits of the local linear versus the Nadaraya---Watson estimator depend very much on the shape of some underlying functions.