Estimation Of Partially Linear Varying Coefficient Transformation Model With Measurement Error

The partially linear varying coefficient transformation model is an extension of the classical proportional hazards model.Because this model is very flexible for data modeling,it has very important applications in many research fields such as biomedicine.This paper studies the estimation equation method of the model,and the estimation equation method is extended to the estimation of the partially linear varying coefficient transformation model with measurement error.In order to estimate the partially linear varying coefficient transformation model.Firstly,B-spline technique is used to approximate the unknown function coefficient.The purpose is to transform the varying coefficient transformation model into the linear transformation model.Next,based on the martingale estimation equation method,the parameters and the B-spline coefficients are estimated to obtain the estimation of the unknown function.Then,based on the calibration estimation method of the verification sample,the estimation equation method is extended to the estimation problem of the linear partial covariate with measurement error.Finally,the validity of the estimation method is verified by numerical simulation.