Population Size Estimation with Covariate Values Missing Non-ignorable  

作　　者：Li-ping LIU Zhi-chao GUO Xiao-gang DUAN 

机构地区：[1]LMAM and CSS, School of Mathematical Sciences, Peking University, Beijing 100871, China [2]Beijing Normal University, Beijing 100875, China

出　　处：《Acta Mathematicae Applicatae Sinica》2016年第3期659-668,共10页应用数学学报（英文版）

基　　金：Supported in part by the National Natural Science Foundation of China under Grant No.11171006

摘　　要：The main purpose of this paper is using capture-recapture data to estimate the population size when some covariate values are missing, possibly non-ignorable. Conditional likelihood method is adopted, with a sub-model describing various missing mechanisms. The derived estimate is proved to be asymptotically normal, and simulation studies via a version of EM algorithm show that it is approximately unbiased. The proposed method is applied to a real example, and the result is compared with previous ones.The main purpose of this paper is using capture-recapture data to estimate the population size when some covariate values are missing, possibly non-ignorable. Conditional likelihood method is adopted, with a sub-model describing various missing mechanisms. The derived estimate is proved to be asymptotically normal, and simulation studies via a version of EM algorithm show that it is approximately unbiased. The proposed method is applied to a real example, and the result is compared with previous ones.

关 键 词：CAPTURE-RECAPTURE conditional likelihood EM algorithm missing non-ignorable 

分 类 号：O212.1[理学—概率论与数理统计]