DOI: http://dx.doi.org/10.18273/revint.v35n1-2017004

Original article

Power Birnbaum-Saunders Student t distribution

Distribución Birnbaum-Saunders Potencia t de Student

 

Germán Moreno-Arenas1

Guillermo Martínez-Flórez2

Heleno Bolfarin3

 

1Universidad Industrial de Santander, Escuela de Matemáticas, Bucaramanga, Colombia.

2Universidad de Córdoba, Departamento de Matemáticas y Estadística, Montería, Colombia.

3Universidade de São Paulo, Departamento de Estatística, São Paulo, Brazil.

 

E-mail: gmorenoa@uis.edu.co

 

Abstract:

The fatigue life distribution proposed by Birnbaum and Saunders has been used quite effectively to model times to failure for materials subject to fatigue. In this article, we introduce an extension of the classical Birnbaum-Saunders distribution substituting the normal distribution by the power Student t distribution. The new distribution is more flexible than the classical Birnbaum-Saunders distribution in terms of asymmetry and kurtosis. We discuss maximum likelihood estimation of the model parameters and associated regression model. Two real data set are analysed and the results reveal that the proposed model better some other models proposed in the literature.

Keywords: Birnbaum-Saunders distribution, alpha-power distribution, power Student t distribution.

 

Resumen.

La distribución de probabilidad propuesta por Birnbaum y Saunders se ha usado con bastante eficacia para modelar tiempos de falla de materiales sujetos a la fátiga. En este artículo definimos una extensión de la distribución Birnbaum-Saunders clásica sustituyendo la distribución normal por la distribución potencia t de Student. La nueva distribución es más flexible que la distribución Birnbaum-Saunders clásica en términos de asimetría y curtosis. Presentamos los estimadores de máxima verosimilitud de los parámetros del modelo y su modelo de regresión asociado. El análisis de dos aplicaciones con datos reales revelan una superioridad del nuevo modelo con  relación a otros modelos existentes en la literatura

Palabras clave: Distribución Birnbaum-Saunders, distribución alfa potencia, distribución potencia t de Student

 

Received: 02 November 2015

Accepted: 26 May 2017.

 

Texto complete disponible en PDF

 References

1. Akaike H., “A new look at statistical model identification”, IEEE Trans. Automat. Control 19 (1974), No. 6, 716–723.

2. Berkane M., Kano Y. and Bentler P.M., “Pseudo maximum likelihood estimation in elliptical theory: effects of misspecification”, Comput. Statist. Data Anal. 18 (1994), No. 2, 255–267.

3. Birnbaum Z.W. and Saunders S.C., “A new family of life distributions”, J. Appl. Probab. 6 (1969), 319–327.

4. Castillo E. and Hadi A., “A method for estimating parameters and quantiles of distributions of continuous random variables”, Comput. Statist. Data Anal. 20 (1995), No. 4, 421–439.

5. Castillo N.O., Gómez H.W. and Bolfarine H., “Epsilon Birnbaum-Saunders distribution family: properties and inference”, Statist. Papers 52 (2011), No. 4, 871–883.

6. Cordeiro G.M. and Lemonte A.J., “The exponentiated generalized Birnbaum-Saunders distribution”, Appl. Math. Comput. 247 (2014), 762–779.

7. Chan P.S., Ng H.K.T., Balakrishnan N. and Zhou Q., “Point and interval estimation for extreme-value regression model under Type-II censoring”, Comput. Statist. Data Anal. 52 (2008), No. 8, 4040–4058.

8. Díaz-García J.A. and Leiva-Sánchez V., “A new family of life distributions based on the elliptically contoured distributions”, J. Statist. Plann. Inference 128 (2005), No. 2, 445–457.

9. Durrans S.R., “Distributions of fractional order statistics in hydrology”, Water Resources Research 28 (1992), No. 6, 1649–1655.

10. Fernández C. and Steel M., “Multivariate Student t regression models: Pitfalls and inference”, Biometrika 86 (1999), No. 1, 153–167.

11. Gómez H., Olivares J. and Bolfarine H., “An extension of the generalized BirnbaumSaunders distribution”, Statist. Probab. Lett. 79 (2009), 331–338.

12. Gupta R.D. and Gupta R.C., “Analyzing skewed data by power normal model”, TEST 17 (2008), No. 1, 197–210.

13. Lange K.L., Little R.J.A. and Taylor J.M.G., “Robust statistical modeling using the t distribution”, J. Amer. Statist. Assoc. 84 (1989), No. 408, 881–896.

14. Leiva V., Ponce G., Marchant C. and Bustos O., “Fatigue statistical distributions useful for modeling diameter and mortality of trees”, Rev. Colombiana Estadíst. 35 (2012), No. 3, 349–370.

15. Lemonte A.J., “A log-Birnbaum-Saunders regression model with asymmetric errors”, J. Stat. Comput. Simul. 82 (2012), 1775–1787.

16. Lemonte A.J., “A new extension of the Birnbaum-Saunders distribution”, Braz. J. Probab. Stat. 27 (2013), 133–149.

17. McCool J.I., “Confidence limits for Weibull regression with censored data”, IEEE Trans. on Reliability 29 (1980), 145–150.

18. Martínez-Flórez G., Bolfarine H. and Gómez H.W., “An alpha-power extension for the Birnbaum-Saunders distribution”, Statistics 48 (2014), No. 4, 896–912.

19. Martínez-Flórez G., Bolfarine H. and Gómez H.W., “The log-linear Birnbaum-Saunders power model”, Methodol. Comput. Appl. Probab. (2016), 1–21.

20. Moreno-Arenas G., Martínez-Flórez G. and Barrera-Causil C., “Proportional hazard Birnbaum-Saunders distribution with application to the survival data analysis”, Rev. Colombiana Estadíst. 39 (2016), No. 1, 129–147.

21. Ng H., Kundu D. and Balakrishnan N., “Modified moment estimation for the two-parameter Birnbaum-Saunders distribution”, Comput. Statist. Data Anal. 43 (2003), No. 3, 283–298.

22. Paula G., Leiva V., Barros M. and Liu S., “Robust statistical modeling using the BirnbaumSaunders t distribution applied to insurance”, Appl. Stoch. Models Bus. Ind. 28 (2012), No. 1, 16–34.

23. R Development Core Team, R: A Language and Environment for Statistical Computing, R Foundation for Statistical Computing, Vienna, Austria, 2015.

24. Rieck J.R. and Nedelman J.R., “A log-linear model for the Birnbaum-Saunders distribution”, Technometrics 33 (1991), No. 1, 51–60.

25. Saunders S.C., “A family of random variables closed under reciprocation”, J. Amer. Statist. Assoc. 69 (1974), No. 346, 533–539.

26. Taylor J. and Verbyla A., “Joint modelling of location and scale parameters of the t distribution”, Stat. Model. 4 (2004), No. 2, 91–112.

27. Vilca-Labra F. and Leiva-Sánchez V., “A new fatigue life model based on the family of skew-elliptical distributions”, Comm. Statist. Theory Methods 35 (2006), No. 2, 229–244.

28. Vuong Q., “Likelihood ratio tests for model selection and non-nested hypotheses”, Econometrica 57 (1989), No. 2, 307–333.

 

To cite this article: G. Moreno-Arenas, G. Martínez-Flórez, H.  Bolfarine, Power Birnbaum-Saunders Student t distribution, Rev. Integr. Temas Mat. 35 (2017), No. 1, 51–70.