In this paper a new approach to local likelihood estimation for censored data is proposed. This method employs the full likelihood and alternates between estimating the baseline hazard and estimating the covariate effect. The proposed methodology incorporates multidimensional data via additive models. Some results regarding inference for the covariate effect are also presented.
No takes yet. Share an insight, caveat, or question.
Gentleman et al. (1991) studied this question.