Penalized Semiparametric Density Estimation Of Right Censored Data

Parametric and nonparametric density estimation has been widely studied in literatue, and semiparametric model can combine the advantages of them. Penalized likelihood method was proposed for density estimation with the smoothing parameter controlling the tradeoff between goodness-of-fit and smoothness of the density. Censored data is common in survival analysis and medical science.In this article we propose a penalized likelihood approach for the semiparametric density model for right censored data. An efficient iterative procedure is proposed for estimation. Approximate generalized maximum likelihood criterion from Bayesian point of view is derived for selecting the smoothing parameter. The statement and the sketched proof of the asymptotic results are also given in the article.The finite sample performance of the proposed estimation approach is evaluated through simulation. Real censored data examples are analyzed to demonstrate the performance of the proposed method.