Title: Partially linear transformation model for HIV data
Authors: Wei Zhao - City University of Hong Kong (China) [presenting]
Alan Wan - City University of Hong Kong (Hong Kong)
Peter Gilbert - University of Washington and Fred Hutchinson Cancer Research Center (United States)
Yong Zhou - Chinese Academy of Sciences (China)
Abstract: Length-biased and right-censored data arises frequently in practice. A partially linear transformation model is considered for length-biased and right-censored data to account for both the linear and nonlinear covariate effects on survival time. We adopt the local nonlinear technique and develop a global and a local unbiased estimating equations for the simultaneous estimation of unknown covariate effects, which are implemented by an iterative computational algorithm. We establish the asymptotic properties of the proposed estimator under several mild conditions and estimate the standard deviation of the proposed estimator via a bootstrap resampling method. The simulation studies have fully demonstrated the good performance of the proposed estimator under finite sample situation. In addition, the proposed method is further illustrated by two HIV data sets to study the relationship between HIV infection and gender.