Comparing and calibrating discrepancy measures for Bayesian model selection

  1. Horra Navarro, Julián de la
  2. Rodríguez Bernal, María Teresa
Revista:
Sort: Statistics and Operations Research Transactions

ISSN: 1696-2281

Año de publicación: 2012

Volumen: 36

Número: 1

Páginas: 69-80

Tipo: Artículo

DIALNET GOOGLE SCHOLAR lock_openAcceso abierto editor

Otras publicaciones en: Sort: Statistics and Operations Research Transactions

Resumen

Different approaches have been considered in the literature for the problem of Bayesian model selection. Recently, a new method was introduced and analysed in De la Horra (2008) by minimizing the posterior expected discrepancy between the set of data and the Bayesian model, where the chi-square discrepancy was used. In this article, several discrepancy measures are considered and compared by simulation, and it is obtained that the chi-square discrepancy is reasonable to use. Then, an easy method for calibrating discrepancies is proposed, and the behaviour of this approach is studied on simulated data. Finally, a set of real data is analysed.

Referencias bibliográficas

  • Aitkin, M. (1991). Posterior Bayes factors (with discussion). Journal of the Royal Statistical Society B, 53, 11–142.
  • Berger, J. O., Bernardo, J. M. and Sun, D. (2009). The formal definition of reference priors. Annals of Statistics, 37, 905–938.
  • Berger, J. O. and Pericchi, L. R. (1996). The intrinsic Bayes factors for model selection and prediction. Journal of the American Statistical Association, 91, 109–122.
  • Bernardo, J. M. (2005). Intrinsic credible regions: an objective Bayesian approach to interval estimation (with discussion). Test, 14, 317–384.
  • Bernardo, J. M. and Rueda, R. (2002). Bayesian hypothesis testing: a reference approach. International Statistical Review, 70, 351–372.
  • Carota, C. Parmigiani, G. and Polson, N. G. (1996). Diagnostic measures of model criticism. Journal of the American Statistical Association, 91, 753–762.
  • De la Horra, J. (2008). Bayesian model selection: Measuring the χ2 discrepancy with the uniform distribution. Communications in Statistics-Theory and Methods, 37, 1412–1424.
  • De la Horra, J. and Rodrı́guez-Bernal, M. T. (2005). Bayesian model selection: a predictive approach with losses based on distances L1 and L2. Statistics & Probability Letters, 71, 257–265.
  • De la Horra, J. and Rodrı́guez-Bernal, M. T. (2006). Prior density selection as a particular case of Bayesian model selection: a predictive approach. Communications in Statistics-Theory and Methods, 35, 1387–1396.
  • Geisser, S. and Eddy, W. F. (1979). A predictive approach to model selection. Journal of the American Statistical Association, 74, 153–160.
  • Gelfand, A. E. (1995). Model determination using sampling-based methods. In: Gilks, W., Richardson, S., Spiegelhalter, D. (Eds.), Markov Chain Monte Carlo in Practice. Chapman & Hall, London, 145– 161.
  • Gelfand, A. E., Dey, D. K. and Chang, H. (1992). Model determination using predictive distributions with implementation via sampling-based methods. In: Bernardo, J. M., Berger, J. O., Dawid, A. P., Smith, A. F. M. (Eds.), Bayesian Statistics 4. Oxford University Press, Oxford, 147–167.
  • Gelfand, A. E. and Ghosh, S. (1998). Model choice: a minimum posterior predictive loss approach. Biometrika, 85, 1–11.
  • Gutiérrez-Peña, E. and Walker, S. G. (2001). A Bayesian predictive approach to model selection. Journal of Statistical Planning and Inference, 93, 259–276.
  • Laud, P. W. and Ibrahim, J. G. (1995). Predictive model selection. Journal of the Royal Statistical Society B, 57, 247–262.
  • McCulloch, R. E. (1989). Local model influence. Journal of the American Statistical Association, 84, 473– 478.
  • O’Hagan, A. (1995). Fractional Bayes factors for model comparison (with discussion). Journal of the Royal Statistical Society B, 57, 99–138.
  • Proschan, F. (1963). Theoretical explanation of observed decreasing failure rate. Technometrics, 5, 375– 383.
  • Silverman, B. W. (1986). Density Estimation for Statistics and Data Analysis. Chapman and Hall, London.
  • San Martini, A. and Spezzaferri, F. (1984). A predictive model selection criterion. Journal of the Royal Statistical Society B, 46, 296–303.
  • Soofi, E. S., Ebrahimi, N. and Habibullah, M. (1995). Information distinguishability with application to the analysis of failure data. Journal of the American Statistical Association, 90, 657–668.
  • Spiegelhalter, D. J. and Smith, A. F. M. (1982). Bayes factors for linear and log-linear models with vague prior information. Journal of the Royal Statistical Society B, 44, 377–387.
  • Trottini, M. and Spezzaferri, F. (2002). A generalized predictive criterion for model selection. Canadian Journal of Statistics, 30, 79–96.
Fuente de los datos: Dialnet