Prior elicitation in multiple change-point models

Gary Koop, Simon M. Potter

Research output: Working paper

Abstract

This paper discusses Bayesian inference in change-point models. Existing approaches involve placing a (possibly hierarchical) prior over a known number of change-points. We show how two popular priors have some potentially undesirable properties (e.g. allocating excessive prior weight to change-points near the end of the sample) and discuss how these properties relate to imposing a fixed number of changepoints in-sample. We develop a new hierarchical approach which allows some of of change-points to occur out-of sample. We show that this prior has desirable properties and handles the case where the number of change-points is unknown. Our hierarchical approach can be shown to nest a wide variety of change-point models, from timevarying parameter models to those with few (or no) breaks. Since our prior is hierarchical, data-based learning about the parameter which controls this variety occurs.
LanguageEnglish
Place of PublicationLeicester
Number of pages29
Publication statusPublished - Sep 2004

Publication series

NameDepartment of Economics Working Papers
PublisherUniversity of Leicester
Volume04/26

Fingerprint

Prior Elicitation
Change-point Model
Change Point
Multiple Models
Hierarchical Prior
Hierarchical Data
Time-varying Parameters
Nest
Bayesian inference
Unknown

Keywords

  • bayesian inference
  • change-point models
  • change-points
  • statistics
  • economics

Cite this

Koop, G., & Potter, S. M. (2004). Prior elicitation in multiple change-point models. (Department of Economics Working Papers; Vol. 04/26). Leicester.
Koop, Gary ; Potter, Simon M. / Prior elicitation in multiple change-point models. Leicester, 2004. (Department of Economics Working Papers).
@techreport{231fc624dfea4ad8a935c344accd46c6,
title = "Prior elicitation in multiple change-point models",
abstract = "This paper discusses Bayesian inference in change-point models. Existing approaches involve placing a (possibly hierarchical) prior over a known number of change-points. We show how two popular priors have some potentially undesirable properties (e.g. allocating excessive prior weight to change-points near the end of the sample) and discuss how these properties relate to imposing a fixed number of changepoints in-sample. We develop a new hierarchical approach which allows some of of change-points to occur out-of sample. We show that this prior has desirable properties and handles the case where the number of change-points is unknown. Our hierarchical approach can be shown to nest a wide variety of change-point models, from timevarying parameter models to those with few (or no) breaks. Since our prior is hierarchical, data-based learning about the parameter which controls this variety occurs.",
keywords = "bayesian inference, change-point models, change-points, statistics, economics",
author = "Gary Koop and Potter, {Simon M.}",
year = "2004",
month = "9",
language = "English",
series = "Department of Economics Working Papers",
publisher = "University of Leicester",
type = "WorkingPaper",
institution = "University of Leicester",

}

Koop, G & Potter, SM 2004 'Prior elicitation in multiple change-point models' Department of Economics Working Papers, vol. 04/26, Leicester.

Prior elicitation in multiple change-point models. / Koop, Gary; Potter, Simon M.

Leicester, 2004. (Department of Economics Working Papers; Vol. 04/26).

Research output: Working paper

TY - UNPB

T1 - Prior elicitation in multiple change-point models

AU - Koop, Gary

AU - Potter, Simon M.

PY - 2004/9

Y1 - 2004/9

N2 - This paper discusses Bayesian inference in change-point models. Existing approaches involve placing a (possibly hierarchical) prior over a known number of change-points. We show how two popular priors have some potentially undesirable properties (e.g. allocating excessive prior weight to change-points near the end of the sample) and discuss how these properties relate to imposing a fixed number of changepoints in-sample. We develop a new hierarchical approach which allows some of of change-points to occur out-of sample. We show that this prior has desirable properties and handles the case where the number of change-points is unknown. Our hierarchical approach can be shown to nest a wide variety of change-point models, from timevarying parameter models to those with few (or no) breaks. Since our prior is hierarchical, data-based learning about the parameter which controls this variety occurs.

AB - This paper discusses Bayesian inference in change-point models. Existing approaches involve placing a (possibly hierarchical) prior over a known number of change-points. We show how two popular priors have some potentially undesirable properties (e.g. allocating excessive prior weight to change-points near the end of the sample) and discuss how these properties relate to imposing a fixed number of changepoints in-sample. We develop a new hierarchical approach which allows some of of change-points to occur out-of sample. We show that this prior has desirable properties and handles the case where the number of change-points is unknown. Our hierarchical approach can be shown to nest a wide variety of change-point models, from timevarying parameter models to those with few (or no) breaks. Since our prior is hierarchical, data-based learning about the parameter which controls this variety occurs.

KW - bayesian inference

KW - change-point models

KW - change-points

KW - statistics

KW - economics

UR - http://www.le.ac.uk/economics/research/RePEc/lec/leecon/dp04-26.pdf

UR - http://www2.le.ac.uk/departments/economics/research/discussion-papers

M3 - Working paper

T3 - Department of Economics Working Papers

BT - Prior elicitation in multiple change-point models

CY - Leicester

ER -

Koop G, Potter SM. Prior elicitation in multiple change-point models. Leicester. 2004 Sep. (Department of Economics Working Papers).