Proof of Theorem 2.2(ii), p. 436 — under (A3), ρ*(μ) = +∞ for every μ ∈ 𝒴 with μ(Ω) ≠ 1
ProvedConvexRiskFn.Dual.conj_eq_top_of_mass_ne_oneconvex-conjugatep2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1risk-functiontranslation-equivariance
Let be a paired system as in §2 in which contains the constant function , and let be proper and translation equivariant (A3): for all . If has total mass , then
Hence every measure in the dual domain of a translation-equivariant risk function has total mass one: the forward direction of Theorem 2.2(ii).
Formalization Note The constant function is for an element assumed equal to the constant function ; (A3) presupposes that constants lie in .
Preamble
import Mathlib import Definitions.Def_ConvexRiskFn_Dual_Setting
Formal statement
namespace ConvexRiskFn.Dual
/-- Proof of Theorem 2.2(ii), p. 436: if `𝒳` contains the constant function `1` and a proper `ρ`
satisfies (A3), then `ρ*(μ) = +∞` for every `μ ∈ 𝒴` with `μ(Ω) ≠ 1`. -/
theorem conj_eq_top_of_mass_ne_one {Ω : Type*} [MeasurableSpace Ω] {𝒳 : Type*} [AddCommGroup 𝒳] [Module ℝ 𝒳]
[TopologicalSpace 𝒳] [IsTopologicalAddGroup 𝒳] [ContinuousSMul ℝ 𝒳] [LocallyConvexSpace ℝ 𝒳]
(S : PairedSpaces Ω 𝒳) (one : 𝒳)
(hone : S.toFun one = fun _ => 1) (ρ : 𝒳 → EReal) (hp : IsProper ρ) (hA3 : A3 one ρ)
(μ : MeasureTheory.SignedMeasure Ω) (hμ : μ ∈ S.Y) (hmass : μ Set.univ ≠ 1) :
conj S ρ μ = ⊤ := by sorry
end ConvexRiskFn.Dual
Source
Ruszczyński, Shapiro, Optimization of convex risk functions, Math. Oper. Res. 31 (2006), p. 436, proof of Theorem 2.2(ii)
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.