The paper describes a new class of capture-recapture models for closed populations when individual covariates are available. The novelty consists in combining a latent class model for capture probabilities where the class weights and the conditional distributions given the latent may depend on covariates, with a model for the marginal distribution of the available covariates as in Liu et al, Biometrika (2017). In addition, the conditional distributions given the latent and covariates are allowed to take into account any general form of serial dependence. A Fisher scoring algorithm for maximum likelihood estimation is presented, and a powerful result based on the implicit function theorem is used to show that the marginal distribution of observed covariates is uniquely determined, once an estimate of the probabilities of being never captured is available. Asymptotic results are outlined, and a procedure for constructing likelihood based confidence intervals for the population total is presented. Two examples with real data are used to illustrate the proposed approach
翻译:本文介绍了在个别共同变量存在的情况下封闭人口的新一类捕捉-抓获模式。新颖之处在于将潜在类别概率模型结合在一起,以便捕捉概率,因为当分类加权数和根据潜在值的有条件分布取决于共同变量时,潜在值可能取决于共同变量,而现有共同变量的边际分布模式,如Liu等人,Biometrika (2017年),此外,允许根据潜在值和共同变量的有条件分布考虑到任何一般形式的序列依赖性。提出了用于最大可能性估算的渔业评分算法,并使用了基于隐含函数的强大结果,以表明所观测到的共变量的边际分布是独特的,一旦得到对概率的估算,就将予以确定;概述了初步结果,并提出了一个程序,用以构建基于人口总数信任期的可能性。使用了两个有实际数据的实例,以说明拟议方法。