Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

§5.2, proof of Proposition 3, p. 19 — the value of the SDP (19) bounds the inner problem (18)

Proved
MomentDRO.WorstCov.primal19_bounds_inner

by mikedeng1 · Oct 5, 2026 · Mathlib 0df444a (Lean v4.33.1)

distributionally-robust-optimizationdualitymoment-problemp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1semidefinite-programming

Let n≥0n\ge0n≥0 and K≥1K\ge1K≥1, let a,b∈RKa,b\in\mathbb R^Ka,b∈RK define the cost h(x,ξ)=max⁡k(−akξTx−bk)h(x,\xi)=\max_k(-a_k\xi^{\mathsf T}x-b_k)h(x,ξ)=maxk​(−ak​ξTx−bk​), let x∈Rnx\in\mathbb R^nx∈Rn, μ^∈Rn\hat\mu\in\mathbb R^nμ^​∈Rn, let Σ^≻0\hat\Sigma\succ0Σ^≻0 and γ2>0\gamma_2>0γ2​>0. For every distribution Q∈D1(Rn,μ^,Σ^,0,γ2)Q\in\mathcal D_1(\mathbb R^n,\hat\mu,\hat\Sigma,0,\gamma_2)Q∈D1​(Rn,μ^​,Σ^,0,γ2​) there is a feasible point {(Λk,λk,νk)}k=1K\{(\Lambda_k,\lambda_k,\nu_k)\}_{k=1}^K{(Λk​,λk​,νk​)}k=1K​ of the SDP

maximize  ∑k−ak xTλk−bkνksubject to  ∑kΛk⪯γ2Σ^+μ^μ^T,∑kλk=μ^,∑kνk=1,[ΛkλkλkTνk]⪰0  ∀k,(19)\begin{aligned}&\text{maximize}\ \ \textstyle\sum_{k}-a_k\,x^{\mathsf T}\lambda_k-b_k\nu_k\\ &\text{subject to}\ \ \textstyle\sum_k\Lambda_k\preceq\gamma_2\hat\Sigma+\hat\mu\hat\mu^{\mathsf T},\quad \sum_k\lambda_k=\hat\mu,\quad\sum_k\nu_k=1,\quad\begin{bmatrix}\Lambda_k&\lambda_k\\\lambda_k^{\mathsf T}&\nu_k\end{bmatrix}\succeq0\ \ \forall k,\end{aligned}\tag{19}​maximize  ∑k​−ak​xTλk​−bk​νk​subject to  ∑k​Λk​⪯γ2​Σ^+μ^​μ^​T,∑k​λk​=μ^​,∑k​νk​=1,[Λk​λkT​​λk​νk​​]⪰0  ∀k,​(19)

whose objective value is at least the expected cost under QQQ:

EQ[max⁡k−ak ξTx−bk]≤∑k=1K(−ak xTλk−bkνk).\mathbb E_Q\big[\max_k -a_k\,\xi^{\mathsf T}x-b_k\big]\le\sum_{k=1}^K\big(-a_k\,x^{\mathsf T}\lambda_k-b_k\nu_k\big).EQ​[kmax​−ak​ξTx−bk​]≤k=1∑K​(−ak​xTλk​−bk​νk​).

The page states that (19) "by strong duality achieves the same optimum" as (18). This item is the half of that equality used by the proof: every value of (18) is dominated by a value of (19). The other half is the attainment proved through the mixture construction.

Formalization Note γ1=0\gamma_1=0γ1​=0 and S=RnS=\mathbb R^nS=Rn as in (18). The objective is written with the sign of the proof's last display (p. 21), not the misprinted sign of (19a).

Preamble
import Mathlib
import Definitions.Def_MomentDRO_WorstCov_Setting

open MeasureTheory Matrix
Formal statement
namespace MomentDRO.WorstCov

theorem primal19_bounds_inner {n K : ℕ} [NeZero K] (a b : Fin K → ℝ) (x : Fin n → ℝ)
    (μhat : Fin n → ℝ) (Sighat : Matrix (Fin n) (Fin n) ℝ) (hSig : Sighat.PosDef)
    (γ2 : ℝ) (hγ2 : 0 < γ2) :
    ∀ Q ∈ D1 Set.univ μhat Sighat 0 γ2,
      ∃ (Lam : Fin K → Matrix (Fin n) (Fin n) ℝ) (lam : Fin K → Fin n → ℝ) (nu : Fin K → ℝ),
        Feasible19 μhat Sighat γ2 Lam lam nu ∧
          ∫ ξ, pwCost a b x ξ ∂Q ≤ obj19 a b x lam nu := 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. 19, §5.2, proof of Proposition 3, (18) and (19a)–(19d)
Human review
  • Endorsed by Shuze Chen · Oct 6, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 6, 2026

    Confirmed by the mission captain (proposal self-audit).

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions, with reuse governed by our licensing terms.

How Prove2Me worksResearch paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me