Improvement over Bayes prediction in small samples in the presence of model uncertainty

Hubert Wong, Bertrand S Clarke

Research output: Contribution to journalArticle

8 Citations (Scopus)

Abstract

In an online prediction context, the authors introduce a new class of mongrel criteria that allow for the weighing of candidate models and the combination of their predictions based both on model-based and empirical measures of their performance. They present simulation results which show that model averaging using the mongrel-derived weights leads, in small samples, to predictions that are more accurate than that obtained by Bayesian weight updating, provided that none of the candidate models is too distant from the data generator.

Original languageEnglish (US)
Pages (from-to)269-283
Number of pages15
JournalCanadian Journal of Statistics
Volume32
Issue number3
DOIs
StatePublished - Jan 1 2004

Fingerprint

Model Uncertainty
Bayes
Small Sample
Prediction
Model Averaging
Empirical Measures
Updating
Generator
Model-based
Model
Small sample
Model uncertainty
Simulation

Keywords

  • Forecasting
  • Model averaging
  • Mongrel risk

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Cite this

Improvement over Bayes prediction in small samples in the presence of model uncertainty. / Wong, Hubert; Clarke, Bertrand S.

In: Canadian Journal of Statistics, Vol. 32, No. 3, 01.01.2004, p. 269-283.

Research output: Contribution to journalArticle

Wong, Hubert ; Clarke, Bertrand S. / Improvement over Bayes prediction in small samples in the presence of model uncertainty. In: Canadian Journal of Statistics. 2004 ; Vol. 32, No. 3. pp. 269-283.
@article{1a6b768d854f436d9cf37d2fb6994fb1,
title = "Improvement over Bayes prediction in small samples in the presence of model uncertainty",
abstract = "In an online prediction context, the authors introduce a new class of mongrel criteria that allow for the weighing of candidate models and the combination of their predictions based both on model-based and empirical measures of their performance. They present simulation results which show that model averaging using the mongrel-derived weights leads, in small samples, to predictions that are more accurate than that obtained by Bayesian weight updating, provided that none of the candidate models is too distant from the data generator.",
keywords = "Forecasting, Model averaging, Mongrel risk",
author = "Hubert Wong and Clarke, {Bertrand S}",
year = "2004",
month = "1",
day = "1",
doi = "10.2307/3315929",
language = "English (US)",
volume = "32",
pages = "269--283",
journal = "Canadian Journal of Statistics",
issn = "0319-5724",
publisher = "Statistical Society of Canada",
number = "3",

}

TY - JOUR

T1 - Improvement over Bayes prediction in small samples in the presence of model uncertainty

AU - Wong, Hubert

AU - Clarke, Bertrand S

PY - 2004/1/1

Y1 - 2004/1/1

N2 - In an online prediction context, the authors introduce a new class of mongrel criteria that allow for the weighing of candidate models and the combination of their predictions based both on model-based and empirical measures of their performance. They present simulation results which show that model averaging using the mongrel-derived weights leads, in small samples, to predictions that are more accurate than that obtained by Bayesian weight updating, provided that none of the candidate models is too distant from the data generator.

AB - In an online prediction context, the authors introduce a new class of mongrel criteria that allow for the weighing of candidate models and the combination of their predictions based both on model-based and empirical measures of their performance. They present simulation results which show that model averaging using the mongrel-derived weights leads, in small samples, to predictions that are more accurate than that obtained by Bayesian weight updating, provided that none of the candidate models is too distant from the data generator.

KW - Forecasting

KW - Model averaging

KW - Mongrel risk

UR - http://www.scopus.com/inward/record.url?scp=2342588384&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=2342588384&partnerID=8YFLogxK

U2 - 10.2307/3315929

DO - 10.2307/3315929

M3 - Article

VL - 32

SP - 269

EP - 283

JO - Canadian Journal of Statistics

JF - Canadian Journal of Statistics

SN - 0319-5724

IS - 3

ER -