Page - 216 - in Differential Geometrical Theory of Statistics
Image of the Page - 216 -
Text of the Page - 216 -
Entropy2016,18, 433
This isanother topologicalnatureof theentropy. Foranotherviewpointsee [16,31].
Our purpose is to show the theory of statistical models has a homological nature in the
categoryFB(Γ,Ξ).
Definition59. Astatisticalmodel for ameasurable set (Ξ,Ω) is couple (M,[p]) formedby anobject of the
category (FB−Γ,Ξ), namely
M=[E,π,M,D]
andasmoothΓ-equivariant randomcohomologyclass
[p]∈H0e(E,R).
Further the to everyfiber p|Ex is aprobabilitydensity.
AComment.
Let (U,Φ×φ)bea local chart of [E,π,M,D]and let x∗∈U.Weset
ΘU=φ(U),
(Ex∗,Ωx∗)=Φ−1[{φ(x∗)}×(Ξ,Ω)].
The lastdefinitionabove says thatweobtain theprobability space
(Ex∗,Ωx∗,[p]).
Thispropertydoesnotdependonthe choiceof the local chart (U,Φ×φ). Thuswecanregard [M,p]asa
special typeofhomologicalmap
FB(Γ,Ξ) M→ [p]∈H0e(E,R).
9.TheModuliSpaceof theStatisticalModels
Wearegoing to faceanothermajoropenproblem. Thechallenge is thesearch foran invariant
whichencodes thepointsof theorbit space
m= M
G .
That iswhatiscalledtheproblemofmodulispace.Thisproblemofmodulispaceisamajorchallenge
inboththedifferentialgeometryandthealgebraicgeometry(seethetheoryofTeichmuller).Theproblem
isratherconfusedlyaddressed in [30]. Subsequently itprovokedcontroversiesandcriticisms.
TheHessianFunctor
Weconsider thecategoryBFwhoseobjectsarepairs{M,B} formedbyamanifoldMequipped
bilinear formsB∈Γ(T∗⊗2M).
InPartAwehavedefinedtheHessiandifferentialoperatorofaKoszulconnection∇, namely
D∇=∇2.
Thoseoperatorsareuseful foraddressingtheproblemofmoduli spaces. Forourpurpose four
categoriesare involved,
(1) ThecategoryLCwhoseobjectsaregaugestructures (M,∇),
(2) ThecategoryGMwhoseobjectsarestatisticalmodels formeasurablesets,
(3) thecategoryBFwhoseobjectsaremanifoldsequippedbilinear forms,
216
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