Abstract
Bayesian methods of model discrimination are discussed in this chapter. Alternative Bayes factors are presented for when improper priors are used and the usual Bayes factor cannot be specified. The concepts of imaginary training sample and minimal training samples and of partial , fractional , intrinsic and posterior Bayes factors are defined. Applications of these concepts to alternative (exponential , gamma , Weibull and lognormal ) distributions and to systems of linear regressions are presented. Simulation results are used to compare the alternative Bayes factors . The Full Bayesian Significance Test (FBST) is also presented, with applications to a linear mixture model .
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsReferences
Aitkin, M.: Posterior Bayes factors (with discussion). J. Roy. Stat. Soc. B 51, 111–142 (1991)
Aitkin, M.: Evidence and the posterior factor. Math. Sci. 17, 15–25 (1992)
Aitkin, M.: Posterior Bayes factor analysis for an exponential regression model. Stat. Comput. 3, 17–22 (1993)
Aitkin, M., Boys, R.J., Chadwick, T.: Bayesian point null hypothesis testing via the posterior likelihood ratio. Stat. Comput. 15, 215–230 (2005)
Araujo, M.I.: Comparison of separate models: a bayesian approach using improper prior distributions, Ph.D. Thesis (Operational Research). COPPE/UFRJ-Federal University of Rio de Janeiro (In Portuguese) (1998)
Araujo, M.I., Fernandes, M., Pereira, B.B.: Alternative procedures do discriminate nonnested multivariate linear regression models. Commun. Stat Theory Methods 34, 2047–2062 (2005)
Araujo, M.I., Pereira, B.B.: Bayes factors to discriminate separate multivariate regression using improper prior (in Portuguese). Rev. Bras. Estatí. 68, 33–50 (2007a)
Araujo, M.I., Pereira, B.B.: A comparison of Bayes factors for separated models: some simulation results. Commun. Stat. Simul. Comput. 36, 297–309 (2007b)
Atkinson, A.C.: Posterior probabilities for choosing a regression model. Biometrika 65, 39–48 (1978)
Barbosa, F.H.: The Brazilian inflaction in the post-war: monetarism versus structuralism. IPEA/INPES (1983) (In Portuguese)
Basu, S.A.: A new look at Bayesian point null hypothesis testing. Sankhya Indian J Stat. Ser. A 58, 292–310 (1996)
Berger, J.O., Pericchi, L.R.: The intrinsic Bayes factor for model selection and prediction. J. Am. Stat. Ass. 91, 109–122 (1996)
Berger, J.O., Mortera, J.: Default Bayes factor for nonnested hypotheses testing. J. Am. Stat. Ass. 94, 542–554 (1999)
Berger, J., Pericchi, L.R.: Objective Bayesian model methods for model selection: introduction and comparison (with discussion). In: Lahir, P. (ed.) IMS Lecture Notes-Monograph Series, vol. 18, pp. 135–207 (2001)
Box, G.E.P., Tiao, G.C.: Multiparameter problems from a Bayesian point of view. Ann. Math. Stat. 36, 1468–1482 (1965)
Cox, D.R.: Tests of separate families of hypotheses. In: Proceedings 4th Berkeley Symposium in Mathematical Statistics and Probability, vol. 1, pp. 105–123 (1961). University of California Press
Cox, D.R.: The role of significance tests. Scand. J. Stat. 4, 49–70 (1977)
Diniz, M., Pereira, C.A.B., Polpo, A., Stern, J.M., Wechsler, S.: Relationship between Bayesian and frequentist significance indices. Int. J. Uncertainty Quantif. 2, 161–172 (2012)
Gelfand, A.E., Dey, D.K.: Bayesian model choice: asymptotic and exact calculations. J. Roy. Stat. Soc. B 56, 501–504 (1994)
Isbicki, R., Fossaluza, V., Hounie, A.G., Nakano, E.Y., Pereira, C.A.B.: Testing allele homogeneity: the problem of nested hypotheses. BMC Genet. 13, 103 (2011)
Kass, R.E., Raftery, A.E.: Bayes factors. J. Am. Stat. Assoc. 90, 773–795 (1995)
Kempthorne, O.: Of what use are tests of significance and tests of hypothesis. Commun. Stat. Theory Methods 8, 763–777 (1976)
Lauretto, M.S., Pereira, C.A.B., Stern, J.M., Zacks, S.: Full Bayesian significance test applied to multivariate normal structure models. Braz. J. Probab. Stat. 17, 147–168 (2003)
Lauretto, M.S., de Faria-Jr, S.P. Pereira, C.A.B., Pereira, B.B., Stern, J.: The problem of separate hypotheses via mixture models. In: Knuth, K.H., Caticha, A., Center, J.L., Jr, Griffin, A., Rodrigues, C.C.: Orgs, Bayesian and maximum entropy methods in science and engineering. In: American Institute of Physics Proceedings, vol. 954, pp. 268–275 (2007)
Lauretto, M., Nakano, F., Pereira, C.A.B., Stern, J.M.: A Straightforward multiallelic significance test for Hardy-Weinberg equilibrium law. Genet. Mol. Biol. 32, 619–625 (2009)
Lempers, F.B.: Posterior probabilities of alternative linear models. University of Rotterdam Press (1971)
Lindley, D.V.: On the presentation of evidence. Math. Sci. 18, 60–63 (1993)
Madruga, M.R., Esteves, L.G., Wechser, S.: On the Bayesianity of Pereira-Stern tests. Test 10, 291–299 (2001)
Madruga, M.R., Pereira, C.A.B., Stern, J.M.: Bayesian evidence test for precise hypotheses. J. Stat. Plan. Inference 117, 185–198 (2003)
Melo, B.A.R.: Bayesian analysis of finite mixture models with censure data and covariates. Ph.D, Qualification Report (in Portugues) (2016)
O’Hagan, A.: Fractional Bayes factor for model comparison (with discussion). J. Roy. Stat. Soc. B 1, 99–138 (1995)
O’Hagan, A.: Properties of intrinsic and fractional Bayes factors. Test 6, 101–108 (1997)
Pereira, C.A.B., Wechsler, S.: On the concept of P-Value. Braz. J. Probab. Stat. 7, 159–177 (1993)
Pereira, C.A.B., Stern, J.M.: Evidence and credibility: full Bayesian significance test for precise hypothesis. Entropy 1, 69–80 (1999)
Pereira, C.A.B., Stern, J.M., Wechsler, S.: Can a significance test be genuinely Bayesian. Bayesian Anal. 3, 79–100 (2008)
Pereira, C.A.B., Polpo, A.: Bayesian statistics with applications to categorical and survival data (in portuguese- Estatística Bayesiana com Aplicações em Dados Categóricos e de Sobrevivência). RBRAS, Campina Grande, 61 p (2014)
Pericchi, L.R.: An alternative to standard Bayesian procedure for discrimination between normal linear models. Biometrika 71, 575–581 (1984)
Pericchi, L.R., Pereira, C.A.B.: Adaptative significance level using optimal decision rule: Balancing by weighting the error probability. Braz. J. Probabil. Stat. 30, 70–90 (2016)
Pericchi, L.R.: Model selection and hypothesis testing based on objective probabilities and Bayes factors. In: Dey, D.K., Rao, C.R., Handbook of Statistics: Bayesian Thinking. Modeling and Computation. North Holland, vol. 25, pp. 115–149 (2005)
Rogatko, A., Slifker, M.J., Babb, J.S.: Hardy-Weinberg equilibrium diagnostics. Theoret. Popul. Biol. 62, 251–257 (2002)
Rust, R.T., Schmittlein, D.C.: A Bayesian cross-validated likelihood method for comparing alternative specifications of quantitative models. Mark. Sci. 4, 20–40 (1985)
Smith, A.F.M., Spiegelhalter, D.J.: Bayes factors and choice criteria for linear models. J. Roy. Stat. Soc. B 42, 213–220 (1980)
Spiegelhalter, D.J., Smith, A.F.M.: Bayes factors for linear and log-linear model with vague prior information. J. Roy. Stat. Soc. B 44, 377–387 (1982)
Stern, J.M., Zacks, S.: Testing the independence of poisson variates under the holgate bivariate distribution: the power of new evidence teste. Stat. Probabil. Lett. 60, 313–320 (2002)
West, M., Harrison, J.: Bayesian Forecasting and Dynamic Models, 2nd edn. Springer, New York (1997)
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
Copyright information
© 2016 The Author(s)
About this chapter
Cite this chapter
Pereira, B., Pereira, C. (2016). Bayesian Methods. In: Model Choice in Nonnested Families. SpringerBriefs in Statistics. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-662-53736-7_3
Download citation
DOI: https://doi.org/10.1007/978-3-662-53736-7_3
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-662-53735-0
Online ISBN: 978-3-662-53736-7
eBook Packages: Mathematics and StatisticsMathematics and Statistics (R0)