Efficient estimation of semiparametric varying-coefficient partially linear transformation model with current status data

Abstract

We consider a varying-coefficient partially linear transformation model with current status data, which extends several semiparametric models for current status data in the literature. Sieve maximum likelihood estimation method is used to obtain an integrated estimate for both the parametric components and nonparametric components in the model, i.e. the linear regression coefficients, the varying-coefficient functions and the baseline survival function. Under some regularity conditions, the proposed parameter estimators are proved to be semiparametrically efficient and asymptotically normal, and the estimators for the nonparametric functions achieve the optimal rate of convergence. Simulation studies assure the theoretical results, and a real data is reanalysed using the proposed method and it yields new findings.