Web-Books
im Austria-Forum
Austria-Forum
Web-Books
Naturwissenschaften
Physik
Differential Geometrical Theory of Statistics
Seite - 193 -
  • Benutzer
  • Version
    • Vollversion
    • Textversion
  • Sprache
    • Deutsch
    • English - Englisch

Seite - 193 - in Differential Geometrical Theory of Statistics

Bild der Seite - 193 -

Bild der Seite - 193 - in Differential Geometrical Theory of Statistics

Text der Seite - 193 -

Entropy2016,18, 433 Proposition9. Weassumeamodel (Θ,P) is regular. For everynonzero realnumberαonehas M(∇α,∇−α)∩M(∇−α,∇α)=M(g,∇−α,∇α)+M(g,∇∗,∇). TheSketchofProof. Theproof isbasedontheshortexact sequences 0→M(g,∇α,∇−α)→M(∇α,∇−α)→S∇2 (Θ)→0, 0→M(g,∇−α,∇α)→M(∇−α,∇α)→S−∇2 (Θ)→0. Letussuppose that theconclusionof the theproposition fails. Thenthere isanonzero2-form B∈S∇α2 (Θ)∩S∇ −α 2 (Θ). (1) IfKer(B)=0, thenboth∇α and∇−α coincidewith theLevi-CivitaconnectionofB. This implies α=0, thiscontradictsourchoiceofα. (2) If Ker(B) = 0 then Ker(B) and Ker(B)⊄ are geodesic for both∇α and∇−α. Thus the pair (Ker(B),Ker(B)⊄)deïŹnesag-orthogonal2-web. Atoneside,by thevirtueof the reduction theoremas in [18]every leafFofKer(B)⊄ inheritsa dualpair (F,gF,∇αF,∇−αF ). Atanotherside,Bgivesrise to theRiemannianstructure (F,B). Furthermoreboth∇αF and∇−αF are torsionfreemetricconnections in (F,B). Therebyonegets ∇αF=∇−αF The last equalityholds if andonly ifα= 0.This contradictsourassumption. Theproposition isproved. Thepropositionabove isaseparationcriterionforα-connections in the followingsense. Forevery nonzerorealnumberα thevectorsubspaceS∇α2 (Θ) is transverse toS∇ −α 2 (Θ) in thevectorspaceS2(Θ) ofsymmetric forms inΘ. 5.4. TheExponentialModels and theHyperbolicity Achallengeisthesearchofacriterionfordecidingwhetheramodel(Θ,P) isanexponentialfamily. That is thechallenge in [22]. Still,nowadays, thisproblemisopen. TheFisher informationofaregularexponentialmodel isaHessianRiemannianmetric.Weare goingtodemonstratethat theconverseisgloballytrue.Ourproof isbasedoncohomologicalarguments. In theAppendixAtothispaperweintroduceanewnumerical invariantrb(Θ,P)whichmeasures howfar frombeinganexponential family isamodel (Θ,P). The invariant rbderives fromtheglobalanalysisofdifferentialoperators Dα=D∇α, Dα=D∇ α . Nowwearegoingtoprovideacohomological characterizationofexponentialmodels. BeforepursuingwerecalladeïŹnition. LetΞj, j :=1,...,mbeasystemofEuclideancoordinate functionsofRm. DeïŹnition40. [18]Anm-dimensional statisticalmodel (Θ,P) is called an exponentialmodel for (Ξ,Ω) if there exist amap Ξ Ο→ [C(Ο),F1(Ο),...,Fm(Ο)]∈Rm+1 193
zurĂŒck zum  Buch Differential Geometrical Theory of Statistics"
Differential Geometrical Theory of Statistics
Titel
Differential Geometrical Theory of Statistics
Autoren
Frédéric Barbaresco
Frank Nielsen
Herausgeber
MDPI
Ort
Basel
Datum
2017
Sprache
englisch
Lizenz
CC BY-NC-ND 4.0
ISBN
978-3-03842-425-3
Abmessungen
17.0 x 24.4 cm
Seiten
476
Schlagwörter
Entropy, Coding Theory, Maximum entropy, Information geometry, Computational Information Geometry, Hessian Geometry, Divergence Geometry, Information topology, Cohomology, Shape Space, Statistical physics, Thermodynamics
Kategorien
Naturwissenschaften Physik
Web-Books
Bibliothek
Datenschutz
Impressum
Austria-Forum
Austria-Forum
Web-Books
Differential Geometrical Theory of Statistics