Previous |  Up |  Next

Article

Title: On limiting towards the boundaries of exponential families (English)
Author: Matúš, František
Language: English
Journal: Kybernetika
ISSN: 0023-5954 (print)
ISSN: 1805-949X (online)
Volume: 51
Issue: 5
Year: 2015
Pages: 725-738
Summary lang: English
.
Category: math
.
Summary: This work studies the standard exponential families of probability measures on Euclidean spaces that have finite supports. In such a family parameterized by means, the mean is supposed to move along a segment inside the convex support towards an endpoint on the boundary of the support. Limit behavior of several quantities related to the exponential family is described explicitly. In particular, the variance functions and information divergences are studied around the boundary. (English)
Keyword: exponential family
Keyword: variance function
Keyword: Kullback–Leibler divergence
Keyword: relative entropy
Keyword: information divergence
Keyword: mean parametrization
Keyword: convex support
MSC: 60A10
MSC: 62B10
MSC: 94A17
idZBL: Zbl 06537776
idMR: MR3445980
DOI: 10.14736/kyb-2015-5-0725
.
Date available: 2015-12-16T18:53:09Z
Last updated: 2018-01-10
Stable URL: http://hdl.handle.net/10338.dmlcz/144738
.
Reference: [1] Ay, N.: An information-geometric approach to a theory of pragmatic structuring..The Annals of Probability 30 (2002), 416-436. Zbl 1010.62007, MR 1894113, 10.1214/aop/1020107773
Reference: [2] Barndorff-Nielsen, O.: Information and Exponential Families in Statistical Theory..Wiley, New York 1978. Zbl 1288.62007, MR 0489333
Reference: [3] Brown, L. D.: Fundamentals of Statistical Exponential Families..Inst. of Math. Statist. Lecture Notes - Monograph Series 9 (1986). Zbl 0685.62002, MR 0882001
Reference: [4] Chentsov, N. N.: Statistical Decision Rules and Optimal Inference..Translations of Mathematical Monographs, Amer. Math. Soc., Providence - Rhode Island 1982 (Russian original: Nauka, Moscow, 1972). Zbl 0484.62008, MR 0645898
Reference: [5] Csiszár, I., Matúš, F.: Closures of exponential families..The Annals of Probability 33 (2005), 582-600. Zbl 1068.60008, MR 2123202, 10.1214/009117904000000766
Reference: [6] Csiszár, I., Matúš, F.: Generalized maximum likelihood estimates for exponential families..Probability Theory and Related Fields 141 (2008), 213-246. Zbl 1133.62039, MR 2372970, 10.1007/s00440-007-0084-z
Reference: [7] Graham, R., Knuth, D., Patashnik, O.: Concrete Mathematics. Second edition..Addison-Wesley, Reading, Massachusetts 1994, p. 446. MR 1397498
Reference: [8] Letac, G.: Lectures on Natural Exponential Families and their Variance Functions..Monografias de Matemática 50, Instituto de Matemática Pura e Aplicada, Rio de Janeiro 1992. Zbl 0983.62501, MR 1182991
Reference: [9] Matúš, F., Ay, N.: On maximization of the information divergence from an exponential family..In: Proc. WUPES'03 (J. Vejnarová, ed.), University of Economics, Prague 2003, pp. 99-204.
Reference: [10] Matúš, F.: Optimality conditions for maximizers of the divergence from an EF..Kybernetika 43 (2007), 731-746. MR 2376334
Reference: [11] Matúš, F.: Divergence from factorizable distributions and matroid representations by partitions..IEEE Trans. Inform. Theory 55 (2009), 5375-5381. MR 2597169, 10.1109/tit.2009.2032806
Reference: [12] F., F.Matúš, Rauh, J.: Maximization of the information divergence from an exponential family and criticality..In: Proc. IEEE ISIT 2011, St. Petersburg 2011, pp. 809-813. 10.1109/isit.2011.6034269
Reference: [13] Montúfar, G., J., J. Rauh, Ay, N.: Maximal information divergence from statistical models defined by neural networks..In: Proc. GSI 2013, Paris 2013, Lecture Notes in Computer Science 8085 (2013), 759-766. Zbl 1322.62060, 10.1007/978-3-642-40020-9_85
Reference: [14] Rauh, J.: Finding the maximizers of the information divergence from an exponential family..IEEE Trans. Inform. Theory 57 (2011), 3236-3247. MR 2817016, 10.1109/tit.2011.2136230
Reference: [15] Rockafellar, R. T.: Convex Analysis..Princeton University Press, 1970. Zbl 1011.49013, MR 0274683, 10.1017/s0013091500010142
.

Files

Files Size Format View
Kybernetika_51-2015-5_1.pdf 381.9Kb application/pdf View/Open
Back to standard record
Partner of
EuDML logo