Web-Books
in the Austria-Forum
Austria-Forum
Web-Books
Naturwissenschaften
Physik
Differential Geometrical Theory of Statistics
Page - 86 -
  • User
  • Version
    • full version
    • text only version
  • Language
    • Deutsch - German
    • English

Page - 86 - in Differential Geometrical Theory of Statistics

Image of the Page - 86 -

Image of the Page - 86 - in Differential Geometrical Theory of Statistics

Text of the Page - 86 -

Entropy2016,18, 386 GL(n)×Rn→Rn (A,x) →Ax (171) ThegroupGL(n) isaLiegroup, isasubgroupof thegeneralaffinegroupGA(n), composedof allpairs (A,υ)whereA∈GL(n)andυ∈Rn, thegroupoperationgivenby: (A1,υ1)(A2,υ2)=(A1A2,A1υ2+υ1) (172) GL(n) isanopensubsetofRn 2 ,andmaybeconsideredasn2-dimensionaldifferentialmanifoldwiththe samedifferentiablestructure thanRn 2 .Multiplicationandinversionare infinitelyoftendifferentiable mappings.Consider thevectorspacegl(n)ofrealn×nmatricesandthecommutatorproduct: gl(n)×gl(n)→ gl(n) (A,B) →AB−BA=[A,B] (173) This isaLieproductmakinggl(n) intoaLiealgebra. Theexponentialmapis thenthemapping definedby: exp:gl(n)→GL(n) A → exp(A)= ∞∑ n=0 An n! (174) RestrictingA tohavepositivedeterminant,oneobtains thepositivegeneralaffinegroupGA+(n) thatacts transitivelyonRnby: ((A,υ) ,x) →Ax+υ (175) IncaseofsymmetricpositivedefinitematricesSym+(n),wecanuse theCholeskydecomposition: R=LLT (176) where L is a lower triangularmatrixwith real andpositive diagonal entries, and LT denotes the transposeofL, todefinethesquarerootofR. Given a positive semidefinite matrix R, according to the spectral theorem, the continuous functional calculus can be applied to obtain a matrix R1/2 such that R1/2 is itself positive and R1/2R1/2=R. TheoperatorR1/2 is theuniquenon-negativesquarerootofR. Nn = {ℵ(μ,Σ)/μ∈Rn,Σ∈Sym+n} the class of regular multivariate normal distributions, whereμ is themeanvectorandΣ is the (symmetricpositivedefinite) covariancematrix, is invariant under the transitiveactionofGA(n). The inducedactionofGA(n)onRn×Sym+n is thengivenby: GA(n)×(Rn×Sym+n)→Rn×Sym+n ((A,υ) ,(μ,Σ)) → (Aμ+υ,AΣAT) (177) and GA(n)×Rn→Rn ((A,υ) ,x) →Ax+υ (178) Asthe isotropygroupof (0, In) isequal toO(n),wecanobserve that: Nn=GA(n)/O(n) (179) Nn is an open subset of the vector spaceTn = {(η,Ω)/η∈Rn,Ω∈Symn} and is adifferentiable manifold,where the tangentspaceatanypointmaybe identifiedwithTn. 86
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