Weighted Estimation Of Covariate Adjustment In Constrained Interval Censored Cox Models Under Casecohort Sampling

In large cohort studies,a sampling method is often needed to make the study both cost-effective and effective,in the Cox model,interval censored data is a common kind of censored data,because many factors will cause interference or pollution to the covariate data,for the case-cohort sampling under the interval censored data Cox model covariate interference situation,the filling of interval censored data and the adjustment of the disturbed covariate,in order to more efficient statistical inference,add constraints in the study,The weighted estimation under covariate adjustment in the Cox model of constrained interval censored data during Case-cohort sampling was obtained.Firstly,the probability filling method is proposed under the real covariate,and then the kernel function is used to smooth the disturbed covariates in the Cox model under Case-cohort sampling,and the interval censored data is filled into the right censored data by the idea of filling,and then the constraint is added for weighted estimation,referred to as constraint weighted estimation,and then the asymptotic nature of the constraint-weighted estimator is proved,and the constraint MM algorithm is used to realize numerical calculation.Finally,numerical simulations of unconstrained estimation,constraint estimation and constraint-weighted estimation are carried out,and the performance of the constraint-weighted estimation method is analyzed.The numerical simulation shows that it is necessary to adjust the disturbed covariates,and reasonable constraints can improve the efficiency of parameter estimation,and for the Cox model of interval censored data with constraints during Case-cohort sampling,the use of weighted estimation after filling in the interval censored data under covariate adjustment has good performance.