Some heuristics of kernel based estimators of ratio functions
Authors:
P. N. Patil a;
M. T. Wells b;
J. S. Marron ab
| Affiliations: | a University of North Carolina at Chapel Hill, |
| b Cornell University, |
DOI:
10.1080/10485259408832611
Publication Frequency:
8 issues per year
Subjects:
Mathematical Economics;
Mathematical Finance;
Medical Statistics;
Statistical Theory & Methods;
Statistics;
Statistics for the Biological Sciences;
Stochastic Models & Processes;
Formats available:
PDF
(English)
View Article:
View Article (PDF)
Abstract
Ratio functions for which nonparametric estimators have been considered include the hazard rate and density under random censoring. One estimation method involves individual estimates of the numerator and denominator. An alternative targets the entire function not the separate pieces. The two estimators are not comparable in terms of Mean Integrated Squared Error. However, the second type is seen to be more natural because it admits an elegant and useful martingale representation, and also is pictorially more attractive.
|
| Keywords: Censored and uncensored samples; cumulative hazard function; hazard rate; kernel estimators; martingale |
| view references (13) |

Download Citation


CiteULike
Del.icio.us
BibSonomy
Connotea