Published
Bayesian Estimation for the Centered Parameterization of the Skew-Normal Distribution
Estimación bayesiana para la parametrización centrada de la distribución normal-asimétrica
DOI:
https://doi.org/10.15446/rce.v40n1.55244Keywords:
Point estimation, prior distribution, Metropolis-Hastings algorithm (en)algoritmo de Metropolis-Hastings, distribuciones a priori, estimación puntual (es)
The skew-normal (SN) distribution is a generalization of the normal distribution, where a shape parameter is added to adopt skewed forms. The SN distribution has some of the properties of a univariate normal distribution, which makes it very attractive from a practical standpoint; however, it presents some inference problems. Specifically, the maximum likelihood estimator for the shape parameter tends to infinity with a positive probability. A new Bayesian approach is proposed in this paper which allows to draw inferences on the parameters of this distribution by using improper prior distributions in the centered parametrization'' for the location and scale parameter and a Beta-type for the shape parameter. Samples from posterior distributions are obtained by using the Metropolis-Hastings algorithm. A simulation study shows that the mode of the posterior distribution appears to be a good estimator in terms of bias and mean squared error. A comparative study with similar proposals for the SN estimation problem was undertaken. Simulation results provide evidence that the proposed method is easier to implement than previous ones. Some applications and comparisons are also included.
La distribución Normal Asimétrica (SN) es una generalización de la distribución normal, incluye un parámetro extra que le permite adoptar formas asimétricas. La distribución SN tiene algunas de las propiedades de la distribución normal univariada, lo que la hace muy atractiva desde el punto de vista práctico; sin embargo presenta algunos problemas de inferencia. Particularmente, el estimador de máxima verosimilitud para el parámetro de forma tiende a infinito con probabilidad positiva. Se propone una solución
Bayesiana que permite hacer inferencia sobre los parámetros de esta distribución asignando distribuciones impropias en la “parametrización centrada” para el parámetro de localidad y el de escala y una distribución tipo Beta para el parámetro de forma. Las muestras de las distribuciones posteriores se obtienen utilizando el algoritmo de Metropolis-Hastings. Un estudio de simulación muestra que la moda de la distribución posterior parece ser un buen estimador, en términos de sesgo y error cuadrado medio. Se presenta también un estudio de simulación donde se compara el procedimiento propuesto contra otros procedimientos. Los resultados de simulación proveen evidencia de que el método propuesto es más fácil de implementar que las metodologías previas. Se incluyen también algunas aplicaciones y comparaciones.
https://doi.org/10.15446/rce.v40n1.55244
1Colegio de Postgraduados, Socioeconomía Estadística e Informática-Estadística, Texcoco, México. PhD. Email: perpdgo@gmail.com
2Colegio de Postgraduados, Socioeconomía Estadística e Informática-Estadística, Texcoco, México. PhD. Email: jvillasr@colpos.mx
3Colegio de Postgraduados, Socioeconomía Estadística e Informática-Estadística, Texcoco, México. PhD. Email: sergiop@colpos.mx
4Colegio de Postgraduados, Socioeconomía Estadística e Informática-Estadística, Texcoco, México. PhD. Email: sjavier@colpos.mx
The skew-normal (SN) distribution is a generalization of the normal distribution, where a shape parameter is added to adopt skewed forms. The SN distribution has some of the properties of a univariate normal distribution, which makes it very attractive from a practical standpoint; however, it presents some inference problems. Specifically, the maximum likelihood estimator for the shape parameter tends to infinity with a positive probability. A new Bayesian approach is proposed in this paper which allows to draw inferences on the parameters of this distribution by using improper prior distributions in the "centered parametrization" for the location and scale parameter and a Beta-type for the shape parameter. Samples from posterior distributions are obtained by using the Metropolis-Hastings algorithm. A simulation study shows that the mode of the posterior distribution appears to be a good estimator in terms of bias and mean squared error. A comparative study with similar proposals for the SN estimation problem was undertaken. Simulation results provide evidence that the proposed method is easier to implement than previous ones. Some applications and comparisons are also included.
Key words: Metropolis-Hastings Algorithm, Point Estimation, Prior Distribution.
La distribución Normal Asimétrica (SN) es una generalización de la distribución normal, incluye un parámetro extra que le permite adoptar formas asimétricas. La distribución SN tiene algunas de las propiedades de la distribución normal univariada, lo que la hace muy atractiva desde el punto de vista práctico; sin embargo presenta algunos problemas de inferencia. Particularmente, el estimador de máxima verosimilitud para el parámetro de forma tiende a infinito con probabilidad positiva. Se propone una solución Bayesiana que permite hacer inferencia sobre los parámetros de esta distribución asignando distribuciones impropias en la "parametrización centrada" para el parámetro de localidad y el de escala y una distribución tipo Beta para el parámetro de forma. Las muestras de las distribuciones posteriores se obtienen utilizando el algoritmo de Metropolis-Hastings. Un estudio de simulación muestra que la moda de la distribución posterior parece ser un buen estimador, en términos de sesgo y error cuadrado medio. Se presenta también un estudio de simulación donde se compara el procedimiento propuesto contra otros procedimientos. Los resultados de simulación proveen evidencia de que el método propuesto es más fácil de implementar que las metodologías previas. Se incluyen también algunas aplicaciones y comparaciones.
Palabras clave: algoritmo de Metropolis-Hastings, distribuciones a priori, estimación puntual.
Texto completo disponible en PDF
References
1. Arellano-Valle, R. B. & Azzalini, A. (2008), 'The centred parametrization for the multivariate skew-normal distribution', Journal of Multivariate Analysis 99, 1362-1382.
2. Azzalini, A. (1985), 'A class of distributions which includes the normal ones', Scandinavian Journal of Statistics 12, 171-178.
3. Azzalini, A. (2016), The R package sn: The Skew-Normal and Skew-t distributions (version 1.4-0), Università di Padova, Italia. *http://azzalini.stat.unipd.it/SN
4. Azzalini, A. & Capitanio, A. (1999), 'Statistical applications of the multivariate skew normal distribution', Journal of the Royal Statistical Society 61, 579-602.
5. Azzalini, A. & Genton, M. G. (2008), 'Robust likelihood methods based on the skew-t and related distributions', International Statistical Review 76, 106-129.
6. Box, G. & Tiao, G. (1973), Bayesian inference in statistical analysis, Addison-Wesley series in behavioral science: quantitative methods, Addison-Wesley Pub. Co..
7. Carlin, B. P. & Louis, T. A. (2000), Bayes and Empirical Bayes Methods for Data Analysis, second edn, Chapman - Hall/CRC, New York.
8. Chib, S. & Greenberg, E. (1995), 'Understanding the Metropolis-Hastings Algorithm', The American Statistician 49(4), 327-335.
9. Chiogna, M. (1998), 'Some results on the scalar skew-normal distribution', Journal of the Italian Statistical Society 7, 1-13.
10. Gelman, A. & Rubin, D. B. (1992), 'Inference from iterative simulation using multiple sequences', Statistical Science 7, 457-511.
11. Liseo, B. & Loperfido, N. (2006), 'A note on reference priors for the scalar skew-normal distribution', Journal of Statistical Planning and Inference 136, 373-389.
12. Metropolis, N., Rosembluth, A. W., Teller, M. & Teller, E. (1953), 'Equations of state calculations by fast computing machines', Journal of Chememical Physics 21, 1087-1092.
13. Pewsey, A. (2000), 'Problems of inference for Azzalini's skew-normal distribution', Journal of Applied Statistics 27, 859-870.
14. R Core Team, (2015), R: A Language and Environment for Statistical Computing, R Foundation for Statistical Computing, Vienna, Austria. *http://www.R-project.org/
15. Sartori, N. (2006), 'Bias prevention of maximum likelihood estimates for scalar skew normal and skew t distributions', Journal of Statistical Planning and Inference 136, 4259-4275.
16. Wiper, M., Girón, F. J. & Pewsey, A. (2008), 'Objective Bayesian inference for the half-normal and half-t distributions', Communications in Statistics-Theory and Methods 37(20), 3165 -3185.
Este artículo se puede citar en LaTeX utilizando la siguiente referencia bibliográfica de BibTeX:
@ARTICLE{RCEv40n1a06,
AUTHOR = {Pérez-Rodríguez, Paulino and Villaseñor, José A. and Pérez, Sergio and Suárez, Javier},
TITLE = {{Bayesian Estimation for the Centered Parameterization of the Skew-Normal Distribution}},
JOURNAL = {Revista Colombiana de Estadística},
YEAR = {2017},
volume = {40},
number = {1},
pages = {123-140}
}
References
Arellano-Valle, R. B. & Azzalini, A. (2008), ‘The centred parametrization for the multivariate skew-normal distribution’, Journal of Multivariate Analysis 99, 1362–1382.
Azzalini, A. (1985), ‘A class of distributions which includes the normal ones’, Scandinavian Journal of Statistics 12, 171–178.
Azzalini, A. (2016), The R package sn: The Skew-Normal and Skew-t distributions (version 1.4-0), Università di Padova, Italia. http://azzalini.stat.unipd.it/SN
Azzalini, A. & Capitanio, A. (1999), ‘Statistical applications of the multivariate skew normal distribution’, Journal of the Royal Statistical Society 61, 579– 602.
Azzalini, A. & Genton, M. G. (2008), ‘Robust likelihood methods based on the skew-t and related distributions’, International Statistical Review 76, 106– 129.
Box, G. & Tiao, G. (1973), Bayesian inference in statistical analysis, Addison-Wesley series in behavioral science: quantitative methods, Addison-Wesley Pub. Co.
Carlin, B. P. & Louis, T. A. (2000), Bayes and Empirical Bayes Methods for Data Analysis, second edn, Chapman - Hall/CRC, New York.
Chib, S. & Greenberg, E. (1995), ‘Understanding the Metropolis-Hastings Algorithm’, The American Statistician 49(4), 327–335.
Chiogna, M. (1998), ‘Some results on the scalar skew-normal distribution’, Journalof the Italian Statistical Society 7, 1–13.
Gelman, A. & Rubin, D. B. (1992), ‘Inference from iterative simulation using multiple sequences’, Statistical Science 7, 457–511.
Liseo, B. & Loperfido, N. (2006), ‘A note on reference priors for the scalar skewnormal
distribution’, Journal of Statistical Planning and Inference 136, 373– 389.
Metropolis, N., Rosembluth, A. W., Teller, M. & Teller, E. (1953), ‘Equations of state calculations by fast computing machines’, Journal of Chememical Physics 21, 1087–1092.
Pewsey, A. (2000), ‘Problems of inference for Azzalini’s skew-normal distribution’, Journal of Applied Statistics 27, 859–870.
R Core Team (2015), R: A Language and Environment for Statistical Computing, R Foundation for Statistical Computing, Vienna, Austria. http://www.R-project.org/
Sartori, N. (2006), ‘Bias prevention of maximum likelihood estimates for scalar skew normal and skew t distributions’, Journal of Statistical Planning and Inference 136, 4259–4275.
Wiper, M., Girón, F. J. & Pewsey, A. (2008), ‘Objective Bayesian inference for the half-normal and half-t distributions’, Communications in Statistics-Theory and Methods 37(20), 3165 –3185.
How to Cite
APA
ACM
ACS
ABNT
Chicago
Harvard
IEEE
MLA
Turabian
Vancouver
Download Citation
CrossRef Cited-by
1. João Victor B. de Freitas, Pascal Bondon, Caio L. N. Azevedo, Valdério A. Reisen, Juvêncio S. Nobre. (2024). Scale mixtures of multivariate centered skew-normal distributions. Statistics and Computing, 34(6) https://doi.org/10.1007/s11222-024-10512-7.
2. O. Nejadseyfi, H. J. M. Geijselaers, A. H. van den Boogaard. (2019). Evaluation and assessment of non-normal output during robust optimization. Structural and Multidisciplinary Optimization, 59(6), p.2063. https://doi.org/10.1007/s00158-018-2173-2.
3. Haroon M. Barakat, Abdallh W. Aboutahoun, Naeema El-kadar. (2019). A New Extended Mixture Skew Normal Distribution, With Applications. Revista Colombiana de Estadística, 42(2), p.167. https://doi.org/10.15446/rce.v42n2.70087.
Dimensions
PlumX
Article abstract page views
Downloads
License
Copyright (c) 2017 Revista Colombiana de Estadística

This work is licensed under a Creative Commons Attribution 4.0 International License.
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).