State estimation for aoristic models

Marie-Colette van Lieshout, Robin Luca Markwitz*

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review

1 Citation (Scopus)
32 Downloads (Pure)

Abstract

Aoristic data can be described by a marked point process in time in which the points cannot be observed directly but are known to lie in observable intervals, the marks. We consider Bayesian state estimation for the latent points when the marks are modeled in terms of an alternating renewal process in equilibrium and the prior is a Markov point process. We derive the posterior distribution, estimate its parameters and present some examples that illustrate the influence of the prior distribution. The model is then used to estimate times of occurrence of interval censored crimes.
Original languageEnglish
Pages (from-to)1068-1089
Number of pages22
JournalScandinavian journal of statistics
Volume50
Issue number3
Early online date4 Oct 2022
DOIs
Publication statusPublished - 1 Sept 2023

Keywords

  • UT-Hybrid-D
  • Aoristic data
  • Criminological data
  • Marked temporal point process
  • Markov chain Monte Carlo methods
  • Markov point process state estimation
  • Alternating renewal process

Fingerprint

Dive into the research topics of 'State estimation for aoristic models'. Together they form a unique fingerprint.

Cite this