Prior elicitation in multiple change-point models

G.M. Koop, S. Potter

Research output: Contribution to journalArticle

17 Citations (Scopus)

Abstract

This article discusses Bayesian inference in change-point models. The main existing approaches treat all change-points equally, a priori, using either a Uniform prior or an informative hierarchical prior. Both approaches assume a known number of change-points. Some undesirable properties of these approaches are discussed. We develop a new Uniform prior that allows some of the change-points to occur out of sample. This prior has desirable properties, can be interpreted as "noninformative," and treats the number of change-points as unknown. Artificial and real data exercises show how these different priors can have a substantial impact on estimation and prediction.
LanguageEnglish
Pages751-772
Number of pages21
JournalInternational Economic Review
Volume50
Issue number3
DOIs
Publication statusPublished - Aug 2009

Fingerprint

Change point
Prior elicitation
Prediction
Bayesian inference
Exercise

Keywords

  • change-point models
  • Bayesian inference
  • Uniform prior

Cite this

Koop, G.M. ; Potter, S. / Prior elicitation in multiple change-point models. In: International Economic Review. 2009 ; Vol. 50, No. 3. pp. 751-772.
@article{40a62d543df0433591dda4b7cdccbbe2,
title = "Prior elicitation in multiple change-point models",
abstract = "This article discusses Bayesian inference in change-point models. The main existing approaches treat all change-points equally, a priori, using either a Uniform prior or an informative hierarchical prior. Both approaches assume a known number of change-points. Some undesirable properties of these approaches are discussed. We develop a new Uniform prior that allows some of the change-points to occur out of sample. This prior has desirable properties, can be interpreted as {"}noninformative,{"} and treats the number of change-points as unknown. Artificial and real data exercises show how these different priors can have a substantial impact on estimation and prediction.",
keywords = "change-point models, Bayesian inference, Uniform prior",
author = "G.M. Koop and S. Potter",
note = "Also available as a working paper (2004): http://personal.strath.ac.uk/gary.koop/kp12rev.pdf",
year = "2009",
month = "8",
doi = "10.1111/j.1468-2354.2009.00547.x",
language = "English",
volume = "50",
pages = "751--772",
journal = "International Economic Review",
issn = "0020-6598",
number = "3",

}

Prior elicitation in multiple change-point models. / Koop, G.M.; Potter, S.

In: International Economic Review, Vol. 50, No. 3, 08.2009, p. 751-772.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Prior elicitation in multiple change-point models

AU - Koop, G.M.

AU - Potter, S.

N1 - Also available as a working paper (2004): http://personal.strath.ac.uk/gary.koop/kp12rev.pdf

PY - 2009/8

Y1 - 2009/8

N2 - This article discusses Bayesian inference in change-point models. The main existing approaches treat all change-points equally, a priori, using either a Uniform prior or an informative hierarchical prior. Both approaches assume a known number of change-points. Some undesirable properties of these approaches are discussed. We develop a new Uniform prior that allows some of the change-points to occur out of sample. This prior has desirable properties, can be interpreted as "noninformative," and treats the number of change-points as unknown. Artificial and real data exercises show how these different priors can have a substantial impact on estimation and prediction.

AB - This article discusses Bayesian inference in change-point models. The main existing approaches treat all change-points equally, a priori, using either a Uniform prior or an informative hierarchical prior. Both approaches assume a known number of change-points. Some undesirable properties of these approaches are discussed. We develop a new Uniform prior that allows some of the change-points to occur out of sample. This prior has desirable properties, can be interpreted as "noninformative," and treats the number of change-points as unknown. Artificial and real data exercises show how these different priors can have a substantial impact on estimation and prediction.

KW - change-point models

KW - Bayesian inference

KW - Uniform prior

UR - http://www3.interscience.wiley.com/journal/118518576/home

UR - http://dx.doi.org/10.1111/j.1468-2354.2009.00547.x

U2 - 10.1111/j.1468-2354.2009.00547.x

DO - 10.1111/j.1468-2354.2009.00547.x

M3 - Article

VL - 50

SP - 751

EP - 772

JO - International Economic Review

T2 - International Economic Review

JF - International Economic Review

SN - 0020-6598

IS - 3

ER -