§5.2, proof of Proposition 3, p. 20 — the law of ξ* = ζ_k̃ lies in D₁(ℝⁿ, µ̂, Σ̂, 0, γ₂) and has largest covariance
ProvedMomentDRO.WorstCov.mixture_mem_D1_tightcovariancedistributionally-robust-optimizationmixture-distributionmoment-uncertaintyp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1
Let , , , , and let be feasible for (19) with (19b) tight,
and with for every . Let be distributions on with finite second moments such that and . Then the mixture , which is the law of for an independent index with , satisfies
So the constructed distribution meets the covariance constraint (1b) with equality: it has the largest covariance the set allows.
Formalization Note The hypothesis is the page's "assuming without loss of generality that all ". The mixture is the measure-level form of .
Preamble
import Mathlib import Definitions.Def_MomentDRO_WorstCov_Setting open MeasureTheory Matrix
Formal statement
namespace MomentDRO.WorstCov
theorem mixture_mem_D1_tight {n K : ℕ} [NeZero K]
(μhat : Fin n → ℝ) (Sighat : Matrix (Fin n) (Fin n) ℝ) (hSig : Sighat.PosDef)
(γ2 : ℝ) (hγ2 : 0 < γ2)
(Lam : Fin K → Matrix (Fin n) (Fin n) ℝ) (lam : Fin K → Fin n → ℝ) (nu : Fin K → ℝ)
(hfeas : Feasible19 μhat Sighat γ2 Lam lam nu)
(htight : ∑ k, Lam k = γ2 • Sighat + Matrix.vecMulVec μhat μhat)
(hpos : ∀ k, 0 < nu k)
(Pk : Fin K → Measure (Fin n → ℝ)) (hPk : ∀ k, MomentDRO.Conf.HasSecondMoments (Pk k))
(hmean : ∀ k, MomentDRO.Conf.meanVec (Pk k) = (nu k)⁻¹ • lam k)
(hsecond : ∀ k, MomentDRO.Conf.secondMomentAbout (Pk k) 0 = (nu k)⁻¹ • Lam k) :
mixture nu Pk ∈ D1 Set.univ μhat Sighat 0 γ2 ∧
MomentDRO.Conf.secondMomentAbout (mixture nu Pk) μhat = γ2 • Sighat := by sorry
end MomentDRO.WorstCov
Source
Delage & Ye, Distributionally robust optimization under moment uncertainty with application to data-driven problems, authors' draft of 20 Feb 2008 (OR 58(3), 2010), p. 20, §5.2, proof of Proposition 3, construction of ξ* = ζ_k̃ and the displays for E[ξ*], E[ξ*ξ*ᵀ]
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.