Wavelet estimation of density and hazard rate for randomly right censored data

Abstract

In this study, different estimators of probability density functions and hazard rates are constructed under randomly right censored data. Nonparametric approaches are adopted under the assumption that the density and hazard rate has no specific parametric form. Some currently available methods of density and hazard rate estimation are compared to a modified approach. It is shown that wavelet estimators are competitive with the other available methods, and that no specific method can be uniquely used for all subdensities.