The Fisher information and exponential families parametrized by a segment of means

• Autorzy: Graczyk P. , Mamane S.
• Języki publikacji: EN

Abstrakty

EN

We consider natural and general exponential families (Q_m)_mϵM on R^d parametrized by the means. We study the submodels (Q_{0m1+(1−0)m2})_{0ϵ[0, 1]} parametrized by a segment in the means domain from the point of view of the Fisher information. Such a parametrization allows for a parsimonious model and is particularly useful in practical situations when hesitating between two parameters m₁ and m₂. The most interesting cases are multivariate Gaussian and Wishart models with matrix parameters.

Słowa kluczowe

EN

Fisher information; efficient estimator; exponential family; multivariate Gaussian distribution; Wishart distribution; parsimony

• Wydawca: Wrocław University of Science and Technology
• Czasopismo: Probability and Mathematical Statistics
• Rocznik: 2015
• Tom: Vol. 35, Fasc. 1
• Strony: 73--90
• Opis fizyczny: Bibliogr. 16 poz.

Twórcy

autor

Graczyk P.

• LAREMA, Université d’Angers, 2 Bd Lavoisier, 49045 Angers Cedex 01, France

autor

Mamane S.

• School of Statistics and Actuarial Science, University of the Witwatersrand
• LAREMA, Université d’Angers, Private Bag x3, Wits 2050, South Africa

Bibliografia

• [1] T. Andersson and P. Händel, IEEE Standard 1057. Cramér-Rao bound and the parsimony principle, IEEE Transactions on Instrumentation and Measurement 55 (1) (2006), pp. 44-53.
• [2] J.-F. Bercher and C. Vignat, On minimum Fisher information distributions with restricted support and fixed variance, Inform. Sci. 179 (2009), pp. 3832-3842.
• [3] G. Casella and R. L. Berger, Statistical Inference, Duxbury Advances Series, 2002.
• [4] J. Faraut and A. Korányi, Analysis on Symmetric Cones, Oxford Math. Monogr., Oxford University Press, New York 1994.
• [5] R. M. Gray, Toeplitz and Circulant Matrices: A review, Foundations and Trends in Communications and Information Theory, Vol. 2, Issue 3, NOW Publishers, 2006, pp. 155-239.
• [6] R. D. Gupta and D. Kundu, On the comparison of Fisher information of the Weibull and GE distributions, J. Statist. Plann. Inference 136 (2006), pp. 3130-3144.
• [7] E. L. Lehmann and G. Casella, Theory of Point Estimation, Springer Texts Statist., Springer, New York 1998.
• [8] E. L. Lehmann and J. P. Romano, Testing Statistical Hypotheses, Springer Texts Statist., Springer, New York 2005.
• [9] G. Letac, Lectures on Natural Exponential Families and Their Variance Functions, Monogr. Mat., Vol. 50, IMPA, Rio de Janeiro 1992.
• [10] G. Letac and M. Casalis, Natural exponential families, in: Continuous Multivariate Distributions. Volume 1: Models and Applications, second edition, S. Klotz, N. Balakrishnan, and N. L. Johnson (Eds.), Wiley Ser. Probab. Stat., 2000.
• [11] G. Letac and H. Massam, The noncentral Wishart as an exponential family, and its moments, J. Multivariate Anal. 99 (7) (2008), pp. 1393-1417.
• [12] G. Letac and H. Massam, Existence and non-existence of the noncentral Wishart distributions, preprint, 2011.
• [13] J. R. Magnus and H. Neudecker, Matrix Differential Calculus with Applications in Statistics and Econometrics, Wiley, 2007.
• [14] J. C. Mason and D. C. Handscomb, Chebyshev Polynomials, Chapman & Hall/CRC, 2003.
• [15] R. J. Muirhead, Aspects of Multivariate Statistical Theory, Wiley, Hoboken, New Jersey, 2005.
• [16] V. Rico-Ramirez, P. A. Quintana-Hernandez, S. Hernandez-Castro, and U. M. Diwekar, Fisher information as a novel tool for process control applications, in: 20th European Symposium on Computer Aided Process Engineering – ESCAPE20, S. Pierucci and G. Buzzi Ferraris (Eds.), Elsevier B. V., 2010.

Uwagi

Opracowanie rekordu w ramach umowy 509/P-DUN/2018 ze środków MNiSW przeznaczonych na działalność upowszechniającą naukę (2019).