Publicado

2022-07-01

FLAT LIKELIHOODS: THREE-PARAMETER WEIBULL MODEL CASE

VEROSIMILITUDES PLANAS: CASO DEL MODELO WEIBULL DE TRES PARÁMETROS

DOI:

https://doi.org/10.15446/rev.fac.cienc.v11n2.98450

Palabras clave:

Flat likelihood function, threshold parameter, embedded models, GEV distribution, likelihood contours, profile likelihood function (en)
Función de verosimilitud plana, parámetro umbral, modelo empotrado, contornos de verosimilitud, función de verosimilitud perfil, Distribución de VEG (es)

Descargas

Autores/as

  • José A. Montoya Universidad de Sonora, México
  • Gudelia Figueroa-Preciado Universidad de Sonora, México
Criticisms of maximum likelihood estimation frequently occur when likelihood function shape becomes flat. Although some research have been done regarding the possible causes of a flat likelihood, more work is needed to expand our knowledge on this subject. In this paper we analyze the origin of Weibull flat likelihoods. In particular, we study the severity of the likelihood flatness by examining the limit behaviour of the relative profile likelihood for the three-parameter Weibull threshold parameter, when this parameter goes to infinity. In the cases discussed here, flat likelihoods are not only related to sample size but also to an embedded model problem. Due to the widespread use of the likelihood function in inferential statistical methods, it is important not only to identify factors that can cause flat likelihoods, but also to study the severity of this flattening, in order to develop or apply ad hoc statistical and computational methods for making inferences.
Críticas a la estimación por máxima verosimilitud ocurren frecuentemente cuando la forma de la función de verosimilitud es plana. Aunque se ha realizado investigación respecto a las posibles causas de una verosimilitud plana, es necesario un mayor trabajo para expandir nuestro conocimiento sobre este tema. En este artículo se analiza el origen de verosimilitudes Weibull planas. En particular, se estudia la severidad de la planura de la verosimilitud a través de examinar el comportamiento del límite de la verosimilitud perfil relativa del parámetro umbral del modelo Weibull de tres parámetros, cuando este parámetro se va a infinito. En los casos que aquí se presentan, las verosimilitudes planas no están solamente relacionadas con el tamaño de la muestra sino también con un problema de modelos empotrados. Dado el amplio uso de la función de verosimilitud en métodos estadísticos inferenciales, es importante no solamente identificar los factores que pueden ocasionar verosimilitudes planas, sino también estudiar lo severo de este aplanamiento, a fin de aplicar métodos estadísticos y computacionales ad hoc, al realizar inferencias.

Referencias

Barnard, G. A. (1967). The use of the likelihood function in statistical practice. Proceedings of the Fifth Berkeley Symposium on Mathematical Statistics and Probability, 1, 27--40.

Barnard, G. A. & Sprott D. A. (1983). Likelihood. In: Kotz S, Johnson NL (eds) Encyclopedia of statistical science, Vol 4. Wiley, New York, pp 639--644.

Barndorff-Nielsen, O. E. & Cox, D. R. (1994). Inference and asymptotics. Chapman & Hall/CRC. Boca Raton. DOI: https://doi.org/10.1007/978-1-4899-3210-5

Berger, J. O., Liseo, B. & Wolpert, R. L. (1999). Integrated likelihood methods for eliminating nuisance parameters. Statistical Science, 14(1), 1-28. DOI: https://doi.org/10.1214/ss/1009211804

Bolívar-Cimé, A., Díaz-Francés, E. & Ortega, J. (2015). Optimality of profile likelihood intervals for quantiles of extreme value distributions: application to environmental disasters. Hydrological Sciences Journal, 60(4), 651-670. DOI: https://doi.org/10.1080/02626667.2014.897405

Breusch, T. S., Robertson, J. C. & Welsh, A. H. (1997). The emperor's new clothes: a critique of the multivariate t regression model. Statistica Neerlandica, 51(3), 269-286. DOI: https://doi.org/10.1111/1467-9574.00055

Catchpole, E. A. & Morgan, B. J. (1997). Detecting parameter redundancy. Biometrika, 84(1), 187-196. DOI: https://doi.org/10.1093/biomet/84.1.187

Cheng, R. C. H. & Iles, T. C. (1990). Embedded models in three-parameter distributions and their estimation. Journal of the Royal Statistical Society. Series B (Methodological), 52(1), 135-149. DOI: https://doi.org/10.1111/j.2517-6161.1990.tb01777.x

Cole, S. R., Chu, H. & Greenland, S. (2013). Maximum likelihood, profile likelihood, and penalized likelihood: a primer. American Journal of Epidemiology, 179(2), 252-260. DOI: https://doi.org/10.1093/aje/kwt245

Cousineau, D., Goodman, V. W. & Shiffrin, R. M. (2002). Extending statistics of extremes to distributions varying in position and scale and the implications for race models. Journal of Mathematical Psychology, 46(4), 431-454. DOI: https://doi.org/10.1006/jmps.2001.1399

De Haan, L. (1990). Fighting the arch-enemy with mathematics. Statistica neerlandica, 44(2), 45-68. DOI: https://doi.org/10.1111/j.1467-9574.1990.tb01526.x

Deng, B., Jiang, D. & Gong, J. (2018). Is a three-parameter Weibull function really necessary for the characterization of the statistical variation of the strength of brittle ceramics?. Journal of the European Ceramic Society, 38(4), 2234-2242. DOI: https://doi.org/10.1016/j.jeurceramsoc.2017.10.017

El Adlouni, S., Ouarda, T. B., Zhang, X., Roy, R. & Bobée, B. (2007). Generalized maximum likelihood estimators for the nonstationary generalized extreme value model. Water Resources Research, 43(3), W03410. DOI: https://doi.org/10.1029/2005WR004545

Elmahdy, E. E. & Aboutahoun, A. W.(2013). A new approach for parameter estimation of finite Weibull mixture distributions for reliability modeling. Applied Mathematical Modelling, 37(4), 1800-1810. DOI: https://doi.org/10.1016/j.apm.2012.04.023

Elmahdy, E. E. (2015). A new approach for Weibull modeling for reliability life data analysis. Applied Mathematics and computation, 250, 708-720. DOI: https://doi.org/10.1016/j.amc.2014.10.036

Farcomeni, A. & Tardella, L. (2012). Identifiability and inferential issues in capture-recapture experiments with heterogeneous detection probabilities. Electronic Journal of Statistics, 6, 2602-2626. DOI: https://doi.org/10.1214/12-EJS758

Frery, A. C., Cribari-Neto, F. & De Souza, M. O. (2004). Analysis of minute features in speckled imagery with maximum likelihood estimation. EURASIP Journal on Advances in Signal Processing, 2004(16), 2476-2491. DOI: https://doi.org/10.1155/S111086570440907X

Ghosh, M., Datta, G. S., Kim, D. & Sweeting, T. J. (2006). Likelihood-based inference for the ratios of regression coefficients in linear models. Annals of the Institute of Statistical Mathematics, 58(3), 457-473. DOI: https://doi.org/10.1007/s10463-005-0027-3

Green, E. J., Roesch, F. A., Smith, A. F. & Strawderman, W. E. (1994). Bayesian Estimation for the Three-Parameter Weibull Distribution with Tree Diameter Data. Biometrics, 50(1), 254-269. DOI: https://doi.org/10.2307/2533217

Harter, H. L. & Moore, A. H. (1966). Local-maximum-likelihood estimation of the parameters of three-parameter lognormal populations from complete and censored samples. Journal of the American Statistical Association, 61(315), 842-851. DOI: https://doi.org/10.1080/01621459.1966.10480911

Hirose, H. & Lai, T. L. (1997). Inference from grouped data in three-parameter Weibull models with applications to breakdown-voltage experiments. Technometrics, 39(2), 199-210. DOI: https://doi.org/10.1080/00401706.1997.10485085

Khan, H. M., Albatineh, A., Alshahrani, S., Jenkins, N. & Ahmed, N. U. (2011). Sensitivity analysis of predictive modeling for responses from the three-parameter Weibull model with a follow-up doubly censored sample of cancer patients. Computational Statistics & Data Analysis, 55(12), 3093-3103. DOI: https://doi.org/10.1016/j.csda.2011.05.017

Kalbfleisch, J. G. (1985). Probability and Statistical Inference, Vol. 2. Springer-Verlag. New York. DOI: https://doi.org/10.1007/978-1-4612-1096-2

Koutsoyiannis, D. (2004). Statistics of extremes and estimation of extreme rainfall: I. Theoretical investigation/Statistiques de valeurs extrêmes et estimation de précipitations extrêmes: I. Recherche théorique. Hydrological Sciences Journal, 49(4). DOI: https://doi.org/10.1623/hysj.49.4.575.54430

Kreutz, C., Raue, A., Kaschek, D. & Timmer, J. (2013). Profile likelihood in systems biology. The FEBS journal, 280(11), 2564-2571. DOI: https://doi.org/10.1111/febs.12276

Li, R. & Sudjianto, A. (2005). Analysis of computer experiments using penalized likelihood in Gaussian Kriging models. Technometrics, 47(2), 111-120. DOI: https://doi.org/10.1198/004017004000000671

Lima, V. M. & Cribari-Neto, F. (2019). Penalized maximum likelihood estimation in the modified extended Weibull distribution. Communications in Statistics-Simulation and Computation, 48(2), 334-349. DOI: https://doi.org/10.1080/03610918.2017.1381735

Liu, S., Wu, H. & Meeker, W. Q. (2015). Understanding and addressing the unbounded ``likelihood'' problem. The American Statistician}, 69(3), 191-200. DOI: https://doi.org/10.1080/00031305.2014.1003968

Martins, E. S. & Stedinger, J. R. (2000). Generalized maximum-likelihood generalized extreme-value quantile estimators for hydrologic data. Water Resources Research, 36(3), 737-744. DOI: https://doi.org/10.1029/1999WR900330

Martins, E. S. & Stedinger, J. R. (2001). Generalized maximum likelihood Pareto-Poisson estimators for partial duration series. Water Resources Research, 37(10), 2551-2557. DOI: https://doi.org/10.1029/2001WR000367

Montoya, J. A. (2008). La verosimilitud perfil en la Inferencia Estadística. Centro de Investigación en Matemáticas, A. C., Guanajuato, Gto., México.

Montoya, J. A., Díaz-Francés, E. & Sprott, D.A. (2009). On a criticism of the profile likelihood function. Statistical Papers, 50(1), 195-202. DOI: https://doi.org/10.1007/s00362-007-0056-5

Murphy, S. A. & Van Der Vaart, A. W. (2000). On profile likelihood. Journal of the American Statistical Association, 95(450), 449-465. DOI: https://doi.org/10.1080/01621459.2000.10474219

Pawitan, Y. (2001). In All Likelihood: Statistical Modelling and Inference Using Likelihood. Oxford University Press. New York.

Pewsey, A. (2000). Problems of inference for Azzalini's skewnormal distribution. Journal of Applied Statistics, 27(7), 859-870. DOI: https://doi.org/10.1080/02664760050120542

Raue, A., Kreutz, C., Maiwald, T., Bachmann, J., Schilling, M., Klingmüller, U. & Timmer, J. (2009). Structural and practical identifiability analysis of partially observed dynamical models by exploiting the profile likelihood. Bioinformatics, 25(15), 1923-1929. DOI: https://doi.org/10.1093/bioinformatics/btp358

Serfling, R. J. (2002). Approximation Theorems of Mathematical Statistics. John Wiley & Sons. New York.

Silva, H. P. T. N. & Peiris, T. S. G.(2017). Statistical modeling of weekly rainfall: a case study in Colombo city in Sri Lanka. Proceedings of the Engineering Research Conference (MERCon), Moratuwa, IEEE, 241-246. DOI: https://doi.org/10.1109/MERCon.2017.7980489

Smith, R. L. & Naylor, J. C. (1987). A comparison of maximum likelihood and Bayesian estimators for the three parameter Weibull distribution. Journal of the Royal Statistical Society, 36(3), 358--369. DOI: https://doi.org/10.2307/2347795

Sprott, D. A. (2000). Statistical inference in science. Springer-Verlag. New York.

Sundberg, R. (2010). Flat and multimodal likelihoods and model lack of fit in curved exponential families. Scandinavian Journal of Statistics, 37(4), 632-643. DOI: https://doi.org/10.1111/j.1467-9469.2010.00703.x

Tsionas, E. G. (2001). Likelihood and Posterior Shapes in Johnson's System. Sankhya: The Indian Journal of Statistics, Series B, 63(1), 3-9.

Tumlinson, S. E. (2015). On the non-existence of maximum likelihood estimates for the extended exponential power distribution and its generalizations. Statistics & Probability Letters, 107, 111-114. DOI: https://doi.org/10.1016/j.spl.2015.08.009

Cómo citar

APA

Montoya, J. A. y Figueroa-Preciado, G. (2022). FLAT LIKELIHOODS: THREE-PARAMETER WEIBULL MODEL CASE. Revista de la Facultad de Ciencias, 11(2), 39–53. https://doi.org/10.15446/rev.fac.cienc.v11n2.98450

ACM

[1]
Montoya, J.A. y Figueroa-Preciado, G. 2022. FLAT LIKELIHOODS: THREE-PARAMETER WEIBULL MODEL CASE. Revista de la Facultad de Ciencias. 11, 2 (ago. 2022), 39–53. DOI:https://doi.org/10.15446/rev.fac.cienc.v11n2.98450.

ACS

(1)
Montoya, J. A.; Figueroa-Preciado, G. FLAT LIKELIHOODS: THREE-PARAMETER WEIBULL MODEL CASE. Rev. Fac. Cienc. 2022, 11, 39-53.

ABNT

MONTOYA, J. A.; FIGUEROA-PRECIADO, G. FLAT LIKELIHOODS: THREE-PARAMETER WEIBULL MODEL CASE. Revista de la Facultad de Ciencias, [S. l.], v. 11, n. 2, p. 39–53, 2022. DOI: 10.15446/rev.fac.cienc.v11n2.98450. Disponível em: https://revistas.unal.edu.co/index.php/rfc/article/view/98450. Acesso em: 30 jul. 2024.

Chicago

Montoya, José A., y Gudelia Figueroa-Preciado. 2022. «FLAT LIKELIHOODS: THREE-PARAMETER WEIBULL MODEL CASE». Revista De La Facultad De Ciencias 11 (2):39-53. https://doi.org/10.15446/rev.fac.cienc.v11n2.98450.

Harvard

Montoya, J. A. y Figueroa-Preciado, G. (2022) «FLAT LIKELIHOODS: THREE-PARAMETER WEIBULL MODEL CASE», Revista de la Facultad de Ciencias, 11(2), pp. 39–53. doi: 10.15446/rev.fac.cienc.v11n2.98450.

IEEE

[1]
J. A. Montoya y G. Figueroa-Preciado, «FLAT LIKELIHOODS: THREE-PARAMETER WEIBULL MODEL CASE», Rev. Fac. Cienc., vol. 11, n.º 2, pp. 39–53, ago. 2022.

MLA

Montoya, J. A., y G. Figueroa-Preciado. «FLAT LIKELIHOODS: THREE-PARAMETER WEIBULL MODEL CASE». Revista de la Facultad de Ciencias, vol. 11, n.º 2, agosto de 2022, pp. 39-53, doi:10.15446/rev.fac.cienc.v11n2.98450.

Turabian

Montoya, José A., y Gudelia Figueroa-Preciado. «FLAT LIKELIHOODS: THREE-PARAMETER WEIBULL MODEL CASE». Revista de la Facultad de Ciencias 11, no. 2 (agosto 4, 2022): 39–53. Accedido julio 30, 2024. https://revistas.unal.edu.co/index.php/rfc/article/view/98450.

Vancouver

1.
Montoya JA, Figueroa-Preciado G. FLAT LIKELIHOODS: THREE-PARAMETER WEIBULL MODEL CASE. Rev. Fac. Cienc. [Internet]. 4 de agosto de 2022 [citado 30 de julio de 2024];11(2):39-53. Disponible en: https://revistas.unal.edu.co/index.php/rfc/article/view/98450

Descargar cita

CrossRef Cited-by

CrossRef citations0

Dimensions

PlumX

Visitas a la página del resumen del artículo

206

Descargas

Los datos de descargas todavía no están disponibles.