METHODOLOGY OF EXTREME VALUES ANALYSIS AND ITS APPLICATION FOR PARAMETER ESTIMATION OF GENERALIZED LINEAR MODELS
DOI:
https://doi.org/10.15588/1607-3274-2016-1-3Keywords:
extreme value theory, generalized linear models, extreme value threshold, maximum likelihood method, Bayesian approach.Abstract
The article deals with methodology of extreme values treatment for building and estimating unknown parameters of generalized linearmodels. As a mathematical tool for carrying out the research the extreme value theory was used that creates one of the directions in
mathematical statistics, and is related to investigating the extreme deviations from the median values in probability distributions. Also, the
methods of approximation statistical data to generalized extreme value distribution, the methods of estimating unknown parameters and
selecting an optimal threshold for extreme value models are discussed. The models of treatment extreme values are constructed which are based on actual statistical data and approach is proposed for their future application for estimating predictive models. The model with generalized Pareto distribution turned out to be acceptable for further use, because it has minimum value of observation error and the best approximation of observed curve to theoretical density function. The comparison of evaluation unknown models’ parameters using method of maximum likelihood and Bayesian approach leads to next conclusion. The Bayesian methods are efficient way to solve the problem of selection the best model, based on the received alternatives set and prior parameters values. In future studies it will be reasonable to consider the application of extreme value analysis to predicted generalized linear models.
References
Coles S. An Introduction to Statistical Modeling of Extreme Values / S. Coles. – London : Springer-Verlag, 2001. – P. 45–104. 2. Smith R. L. An overview of Extreme value theory / R. L. Smith. – Lausanne : Bernoulli Center, 2009. 3. Mallor F. An introduction to statistical modeling of extreme values / F. Mallor, E. Nualart, E. Omey // Hub research paper. – 2009. – No. 36. – P. 5–31. 4. Shumway R. H. Time series analysis and its applications / R. H. Shumway, D. S. Stoffer. – New York : Springer, 2006. – 598 p. 5. Romano A. Dynamic learning methods / A. Romano, G. Secundo – New York: Springer, 2009. – 190 p. 6. McCullagh P. Generalized Linear Models / P. McCullagh, J. Nelder.– New York : Chapman & Hall, 1989. – 526 р. 7. Tsay R. S. Analysis of financial time series / R. S. Tsay. – New Jersey : John Wiley & Sons, Inc., 2010. – 715 p. 8. Besag J. Markov Chain Monte Carlo for Statistical Inference / J. Besag. – Center for Statistics and the Social Sciences. – 2001. – No. 9. – 25 p. 9. Бідюк П. І. Оцінювання узагальнених лінійних моделей за байєсівським підходом в актуарному моделюванні / П. І. Бідюк, С. В. Трухан // Наукові Вісті НТУУ «КПІ». – 2014. – № 6. – С.49–55. 10. Beirlant J. Statistics of extremes: Theory and application / J. Beirlant. – New York : John Wiley & Sons, Inc., 2004. – 505 p.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2016 S. B. Trukhan, P. I. Bidyuk
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
Creative Commons Licensing Notifications in the Copyright Notices
The journal allows the authors to hold the copyright without restrictions and to retain publishing rights without restrictions.
The journal allows readers to read, download, copy, distribute, print, search, or link to the full texts of its articles.
The journal allows to reuse and remixing of its content, in accordance with a Creative Commons license СС BY -SA.
Authors who publish with this journal agree to the following terms:
-
Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License CC BY-SA 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.
