Page - 10 - in Differential Geometrical Theory of Statistics
Image of the Page - 10 -
Text of the Page - 10 -
Entropy2016,18, 370
whereπN : T∗N → N is the canonical projectionof the cotangent bundle and J : T∗N →G∗ is a
smoothmapwhoserestriction toeachfibreT∗xNof thecotangentbundle is linear, andis the transpose
of themapX →ϕ(x,X)=ψ(X)(x).
Remark4. Thehomomorphismψof theLie algebraG into theLie algebraA1(N)of smoothvectorfieldsonN
isanactionof thatLiealgebra, in thesensedefinedbelowDefinition11. Thatactioncanbecanonically lifted into
aHamiltonianactionofG onT∗N,endowedwith its canonical symplectic formdθN Definition13. Themap J
is in fact aHamiltonianmomentummap for thatHamiltonianactionProposition5.
LetLL=dvertL :TN→T∗NbetheLegendremapdefinedinDefinition1.
Theorem3 (Euler–PoincaréEquation). With theabovedefinednotations, letγ : [t0,t1]→Nbeasmooth
parametrizedcurve inNandV : [t0,t1]→G bea smoothparametrizedcurve such that, for each t∈ [t0,t1],
ψ (
V(t) )(
γ(t) )
= dγ(t)
dt . (3)
Thecurveγ is apossiblemotionof theLagrangiansystemif andonly ifV satisfies the
equation(
d
dt −ad∗V(t) )(
J◦LL◦ϕ (
γ(t),V(t) ))− J◦d1L(γ(t),V(t))=0. (4)
The interestedreaderwillfindaproofof that theoremin local coordinates in theoriginalNoteby
Poincaré [29].More intrinsicproofscanbefoundin[12,30].Anotherproof ispossible, inwhichthat
theoremisdeducedfromtheEuler-CartanTheorem1.
Remark5. Equation (3) is called the compatibility conditionandEquation (4) is theEuler–Poincaré equation.
It canbewrittenunder the equivalent
form(
d
dt −ad∗V(t) )(
d2L (
γ(t),V(t) ))− J◦d1L(γ(t),V(t))=0. (5)
Examplesofapplicationsof theEuler–Poincaréequationcanbefoundin[5,6,12,30]and, foran
application in thermodynamics, [31].
4.TheHamiltonianFormalism
TheLagrangian formalism can be applied to any smooth Lagrangian. Its application yields
secondorderdifferential equationsonR×N (in local coordinates, theEuler–Lagrange equations)which in
generalarenot solvedwith respect to the secondorderderivatives of theunknown functionswith respect to
time. Theclassicalexistenceandunicity theoremsfor thesolutionsofdifferentialequations (suchas
theCauchy-Lipschitz theorem) therefore cannotbeapplied to these equations.
Under theadditional assumption that theLagrangian ishyper-regular, averyclever changeof
variablesdiscoveredbyWilliamRowanHamilton (LagrangeobtainedhoweverHamilton’sequations
beforeHamilton, butonly inaspecial case, for theslow“variationsofconstants”suchas theorbital
parametersofplanets in thesolarsystem[32,33]).Hamilton[21,22]allowsanewformulationof these
equations in the frameworkof symplecticgeometry. TheHamiltonian formalismdiscussedbelowis the
useof thesenewequations. Itwas latergeneralized independentlyof theLagrangianformalism.
10
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