Abstract
The scale mixtures of Normal distributions are used as a robust alternative to the normal distribution in linear regression modelling, and a non-iterative Bayesian sampling algorithm is developed to obtain independently and identically distributed samples approximately from the observed posterior distributions, which eliminates the convergence problems in iterative Gibbs sampling. Model selection and influential analysis are conducted to choose the best fitted model and to detect the latent outliers. The performances of the methodologies are illustrated through several simulation studies by comparison with the Normal regression and Gibbs sampling, and finally, the US treasury bond prices data is analyzed using the proposed algorithm.
Similar content being viewed by others
References
Abanto-Valle, C., Bandyopadhyay, D., Lachos, V., & Enriquez, I. (2010). Robust Bayesian analysis of heavy-tailed stochastic volatility models using scale mixtures of normal distributions. Computational Statistics and Data Analysis, 12, 2883–2898.
Andrews, D., & Mallows, C. (1974). Scale mixtures of normal distributions. Journal of the Royal Statistical Society Series B, 36(1), 99–102.
Dempster, A., Laird, N., & Rubin, D. (1977). Maximum likelihood from incomplete data via the em algorithm. Journal of the Royal Statistical Society Series B, 39(1), 1–38.
Dempster, A., Laird, N.M., Rubin, D.B. (1980) Iteratively reweighted least squares for linear regression when errors are Normal/independent distributed. In: Multivariate analysis, vol. V, pp. 35–37.
Fernndez, C., & Steel, M. (2000). Bayesian regression analysis with scale mixtures of normals. Econometric Theory, 16(1), 80–101.
Garay, A., Lachos, V., Bolfarine, H., & Cabral, C. (2015a). Bayesian analysis of censored linear regression models with scale mixtures of normal distributions. Journal of Applied Statistics, 42, 2694–2714.
Garay, A., Lachos, V., Bolfarine, H., & Cabral, C. (2015b) Linear censored regression models with scale mixtures of normal distributions. Statistical Papers, 1–32. doi:10.1007/s00362-015-0696-9.
Lachos, V., Cabral, C., & Abanto-Valle, C. (2012). A non-iterative sampling Bayesian method for linear mixed models with normal independent distributions. Journal of Applied Statistics, 39(3), 531–549.
Lange, K., & Sinsheimer, J. (1993). Normal/independent distributions and their applications in robust regression. Journal of Computational and Graphical Statistics, 2(2), 175–198.
Rosa, G., Gianola, D., & Padovani, C. (2004). Bayesian longitudinal data analysis with mixed models and thick-tailed distributions using mcmc. Journal of Applied Statistics, 31(7), 855–873.
Rosa, G., Padovani, C., & Gianola, D. (2003). Robust linear mixed models with normal/independent distributions and Bayesian mcmc implementation. Biometrical Journal, 45(5), 573–590.
Sheather, S. (2009). A modern approach to regression with R. New York: Springer.
Siegel, A. (1977). Practical business statistics. Boston: Irwin McGraw-Hill.
Tan, M., Tian, G., & Ng, K. (2003). A non-iterative sampling method for computing posteriors in the structure of em-type algorithms. Statistica Sinica, 13(3), 625–640.
Tan, M., Tian, G., & Ng, K. (2010). Bayesian missing data problems: EM, data augmentation and noniterative computation. biostatistics series. New York: Chapman & Hall/CRC.
Zellner, A. (1976). Bayesian and non-Bayesian analysis of the regression model with multivariate Student-t error terms. Journal of the American Statistical Association, 71(354), 400–405.
Acknowledgments
The authors gratefully acknowledge the editor and referees for their valuable comments and suggestions. This research is supported by The National Science Foundation of China Grants 11371227.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Yang, F., Yuan, H. A Non-iterative Bayesian Sampling Algorithm for Linear Regression Models with Scale Mixtures of Normal Distributions. Comput Econ 49, 579–597 (2017). https://doi.org/10.1007/s10614-016-9580-5
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10614-016-9580-5