Seite - 16 - in Differential Geometrical Theory of Statistics
Bild der Seite - 16 -
Text der Seite - 16 -
Entropy2016,18, 370
Thisbracket still satifies theproperties stated inDefinition6, thereforedefinesonG∗ aPoissonstructure
called its canonicalLie-PoissonstructuremodifiedbyΘ.
4.4.2. PropertiesofPoissonManifolds
The interestedreaderwillfindtheproofsof theproperties recalledhere in [8–11].
1. On aPoissonmanifoldP, the Poisson bracket {f,g} of two smooth functions fand g can be
expressedbymeansofasmoothfieldofbivectorsΛ:
{f,g}=Λ(df,dg) , f andg∈C∞(P,R) ,
calledthePoissonbivectorfieldofP. TheconsideredPoissonmanifold isoftendenotedby (P,Λ).
ThePoissonbivectorfieldΛ identicallysatisfies
[Λ,Λ]=0,
the bracket [ , ] in the left hand side being theSchouten-Nijenhuis bracket. That bivector field
determinesavectorbundlemorphismΛ :T∗P→TP,definedby
Λ(η,ζ)= 〈
ζ,Λ (η) 〉
,
whereηandζ∈T∗Pare twocovectorsattachedto thesamepoint inP.
Readers interested in the Schouten-Nijenhuis bracketwill find thoroughpresentations of its
properties in [40,41].
2. Let (P,Λ)beaPoissonmanifold.A(maybetime-dependent)vectorfieldonPcanbeassociated
toeach (maybe time-dependent) smooth functionH :R×P→R. It is called theHamiltonian
vectorfieldassociatedto theHamiltonianH, anddenotedbyXH. Itsexpression is
XH(t,x)=Λ (x) (
dHt(x) )
,
where dHt(x) = dH(t,x)− ∂H(t,x)
∂t dt is the differential of the function deduced from H by
considering tasaparameterwithrespect towhichnodifferentiation ismade.
TheHamiltonequationdeterminedbythe (maybetime-dependent)HamiltonianH is
dϕ(t)
dt =XH( (
t,ϕ(t) )
=Λ (dHt) (
ϕ(t) )
.
3. AnyPoissonmanifold is foliated,byageneralizedfoliationwhose leavesmaynotbeallof the
samedimension, into immersedconnectedsymplecticmanifoldscalledthe symplectic leavesof
thePoissonmanifold. The value, at anypoint of a Poissonmanifold, of the Poissonbracket
of twosmooth functionsonlydependson the restrictionsof these functions to the symplectic
leaf through the consideredpoint, and can be calculated as the Poisson bracket of functions
definedonthat leaf,with thePoissonstructureassociatedto thesymplectic structureof that leaf.
ThispropertywasdiscoveredbyAlanWeinstein, inhisvery thoroughstudyof the local structure
ofPoissonmanifolds [42].
5.HamiltonianSymmetries
5.1. Presymplectic,Symplectic andPoissonMapsandVectorFields
LetMbeamanifoldendowedwithsomestructure,whichcanbeeither
• a presymplectic structure, determinedbyapresymplectic form, i.e., a 2-formω which is closed
(dω=0),
16
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