Capture-Recapture Models and Bayesian Sampling

Abstract

Capture-recapture models are widely used to estimate the unknown size of a closed population, N. A successful strategy for exploiting information about N in this setting is obtained through Bayesian modelling, as shown in Castledine (1981). However, direct Bayesian approaches are often cumbersome to implement in this setting. In this paper, we show how Bayesian sampling, using Gibbs sampling and data augmentation, is particularly well suited for use in a wide variety of capture-recapture models, including the multinomial and classical hypergeometric models. This approach can provide accurate approximations of posterior expressions, including the entire posterior distribution.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Sep 11, 1990
Accession Number
ADA226853

Entities

People

  • Christian P. Robert
  • Edward I. George

Organizations

  • Stanford University

Tags

DTIC Thesaurus Topics

  • Bayesian Inference
  • Bayesian Networks
  • Classification
  • Computational Science
  • Data Science
  • Estimators
  • Information Processing
  • Information Science
  • Military Research
  • Monte Carlo Method
  • Probability
  • Sampling
  • Simulations
  • Statistical Analysis
  • Statistical Samples
  • Statistics
  • Surveys

Fields of Study

  • Mathematics

Readers

  • Marine Hydrodynamics
  • Statistical inference.

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference
  • AI & ML - Machine Learning Algorithms