In many real epidemiology or clinical trials, doubly censoring is a common practice. Where the generated data sets may result in right or left censored failure times along with complete times. In this article, the nonparametric maximum likelihood estimation technique for approximating the survival function when some covariates are involved under doubly censoring scheme is employed. The Taylor series is used to extract the baseline hazard function in the Cox model and hence the likelihood ratio test is also used to determine the appropriate order for Taylor series. This analytical approach demonstrates by a simulation study followed by a real case study using HIV data set.
A. Aljawadi, Bader
"Taylor Series for Survival Function Approximation using Doubly Censoring with Covariates,"
Journal of Statistics Applications & Probability: Vol. 4:
3, Article 2.
Available at: https://digitalcommons.aaru.edu.jo/jsap/vol4/iss3/2