Web-Books
in the Austria-Forum
Austria-Forum
Web-Books
Naturwissenschaften
Physik
Differential Geometrical Theory of Statistics
Page - 284 -
  • User
  • Version
    • full version
    • text only version
  • Language
    • Deutsch - German
    • English

Page - 284 - in Differential Geometrical Theory of Statistics

Image of the Page - 284 -

Image of the Page - 284 - in Differential Geometrical Theory of Statistics

Text of the Page - 284 -

Entropy2016,18, 407 Proof. Forα=±1,equality (39) followstrivially. Thus,weassumeα∈ (−1,1). By (34),wecanwrite ( ∂ ∂θi ) pθ ( ∂ ∂θk ) pϑ κ(α)=−1+α 2 ∫ T[ 1−α 2 ∂ϕ−1(pθ) ∂θi +( ∂ ∂θi )pθκ(α)u0] ∂ϕ−1(pϑ) ∂θk ϕ′′(cα)dμ∫ Tu0ϕ ′(cα)dμ + 1+α 2 ∫ T ∂ϕ−1(pϑ) ∂θk ϕ′(cα)dμ∫ Tu0ϕ ′(cα)dμ ∫ Tu0[ 1−α 2 ∂ϕ−1(pθ) ∂θi +( ∂ ∂θi )pθκ(α)u0]ϕ ′′(cα)dμ∫ Tu0ϕ ′(cα)dμ . (40) Applying ( ∂ ∂θj )pθ to the first termon the right-hand side of (40), and then equating pϑ = pθ, weobtain − 1−α 2 4 E′′θ [∂2ϕ−1(pθ) ∂θi∂θj ∂ϕ−1(pθ) ∂θk ] − 1+α 2 ( ∂2 ∂θi∂θj ) pθ κ(α)E′′θ [ u0 ∂ϕ−1(pθ) ∂θk ] − 1−α 2 4 1−α 2 E′′′θ [∂ϕ−1(pθ) ∂θi ∂ϕ−1(pθ) ∂θj ∂ϕ−1(pθ) ∂θk ] + 1−α2 4 1−α 2 E′′θ [∂ϕ−1(pθ) ∂θi ∂ϕ−1(pϑ) ∂θk ] E′′θ [ u0 ∂ϕ−1(pϑ) ∂θj ] . (41) Similarly, ifweapply ( ∂ ∂θj )pθ to thesecondtermontheright-handsideof (40), andmake pϑ= pθ, weget 1−α2 4 1−α 2 E′′θ [∂ϕ−1(pθ) ∂θj ∂ϕ−1(pθ) ∂θk ] E′′θ [ u0 ∂ϕ−1(pθ) ∂θi ] . (42) Collecting(41)and(42),wecanwrite Γ(α)ijk =− 4 1−α2 [( ∂2 ∂θi∂θj ) pθ ( ∂ ∂θk ) pϑ κ(α) ] pθ=pϑ =E′′θ [∂2ϕ−1(pθ) ∂θi∂θj ∂ϕ−1(pθ) ∂θk ] + 1−α 2 E′′′θ [∂ϕ−1(pθ) ∂θi ∂ϕ−1(pθ) ∂θj ∂ϕ−1(pθ) ∂θk ] − 1−α 2 E′′θ [∂ϕ−1(pθ) ∂θj ∂ϕ−1(pθ) ∂θk ] E′′θ [ u0 ∂ϕ−1(pθ) ∂θi ] − 1−α 2 E′′θ [∂ϕ−1(pθ) ∂θi ∂ϕ−1(pϑ) ∂θk ] E′′θ [ u0 ∂ϕ−1(pϑ) ∂θj ] − 1+α 2 E′θ [∂2ϕ−1(pθ) ∂θi∂θj ] E′′θ [ u0 ∂ϕ−1(pθ) ∂θk ] , (43) whereweused ( ∂2 ∂θi∂θj ) pθ κ(α)= 1−α2 4 [( ∂2 ∂θi∂θj ) pθ D(α)ϕ (pθ ‖ pϑ) ] pϑ=pθ = 1−α2 4 gij=−1−α 2 4 E′θ [∂2ϕ−1(pθ) ∂θi∂θi ] . Expression(39) followsfrom(37), (38)and(43). 5.Conclusions In [17,18], the authors introduced a pair of dual connections D(−1) and D(1) induced by ϕ-divergence. The main motivation of the present work was to find a (non-trivial) family of α-divergences,whose inducedα-connectionsareconvexcombinationsofD(−1) andD(1). Asaresult ofourefforts,weproposedageneralizationofRényidivergence. TheconnectionD(α) inducedbythe generalizationofRényidivergencesatisfies the relationD(α) = 1−α2 D (−1)+ 1+α2 D (1). Togeneralize Rényidivergence,wemadeuseofpropertiesofϕ-functions. Thismakesevident the importanceof ϕ-functions in thegeometryofnon-standardmodels. Instandardstatisticalmanifolds,eventhough Amari’sα-divergenceandRényidivergence (withα∈ [−1,1])donotcoincide, they induce thesame familyofα-connections. Thisstrikingresult requires further investigation. Futureworkshouldfocus 284
back to the  book Differential Geometrical Theory of Statistics"
Differential Geometrical Theory of Statistics
Title
Differential Geometrical Theory of Statistics
Authors
Frédéric Barbaresco
Frank Nielsen
Editor
MDPI
Location
Basel
Date
2017
Language
English
License
CC BY-NC-ND 4.0
ISBN
978-3-03842-425-3
Size
17.0 x 24.4 cm
Pages
476
Keywords
Entropy, Coding Theory, Maximum entropy, Information geometry, Computational Information Geometry, Hessian Geometry, Divergence Geometry, Information topology, Cohomology, Shape Space, Statistical physics, Thermodynamics
Categories
Naturwissenschaften Physik
Web-Books
Library
Privacy
Imprint
Austria-Forum
Austria-Forum
Web-Books
Differential Geometrical Theory of Statistics