Penalized log-likelihood estimation for partly linear transformation models with current status data

摘要

We consider partly linear transformation models applied to current statusdata. The unknown quantities are the transformation function, a linearregression parameter and a nonparametric regression effect. It is shown thatthe penalized MLE for the regression parameter is asymptotically normal andefficient and converges at the parametric rate, although the penalized MLE forthe transformation function and nonparametric regression effect are only$n^{1/3}$ consistent. Inference for the regression parameter based on a blockjackknife is investigated. We also study computational issues and demonstratethe proposed methodology with a simulation study. The transformation models andpartly linear regression terms, coupled with new estimation and inferencetechniques, provide flexible alternatives to the Cox model for current statusdata analysis.