Please use this identifier to cite or link to this item:

Global Partial Likelihood for Nonparametric Proportional Hazards Models

Authors Chen, Kani View this author's profile
Guo, Shaojun
Sun, Liuquan
Wang, Jane-Ling
Issue Date 2010
Source Journal of the American Statistical Association , v. 105, (490), 2010, JUN, p. 750-760
Summary As an alternative to the local partial likelihood method of Tibshirani and Hastie and Fan, Gijbels, and King. a global partial likelihood method is proposed to estimate the covariate effect in a nonparametric proportional hazards model, lambda(t vertical bar x) = exp{psi(x)}lambda(0)(t). The estimator, (psi) over cap (x), reduces to the Cox partial likelihood estimator if the covariate is discrete. The estimator is shown to be consistent and semiparametrically efficient for linear functionals of psi(x). Moreover, Breslow-type estimation of the cumulative baseline hazard function, using the proposed estimator (psi) over cap (x), is proved to be efficient. The asymptotic bias and variance are derived under regularity conditions. Computation of the estimator involves an iterative but simple algorithm. Extensive simulation studies provide evidence supporting the theory. The method is illustrated with the Stanford heart transplant data set. The proposed global approach is also extended to a partially linear proportional hazards model and found to provide efficient estimation of the slope parameter. This article has the supplementary materials online.
ISSN 0162-1459
Language English
Format Article
Access View full-text via DOI
View full-text via Web of Science
View full-text via Scopus