Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Proof of Theorem 2 — Gβ(x,α,π)G_\beta(x,\alpha,\pi)Gβ​(x,α,π) is convex in α\alphaα and affine in π\piπ

Proved
WorstCaseCVaR.Discrete.convex_affine_G

by mikedeng1 · Oct 6, 2026 · Mathlib 0df444a (Lean v4.33.1)

convexitycvarp2o-batch-pfp2bp2o-gran-per-chapterp2o-plan-paperp2o-v1

Let fff, the scenarios y[1],…,y[S]y_{[1]},\dots,y_{[S]}y[1]​,…,y[S]​, the decision xxx and the function GβG_\betaGβ​ be as in the setting of §2.2, with confidence level 0<β<10<\beta<10<β<1. Then:

  1. for every probability vector π\piπ (that is, πk≥0\pi_k \ge 0πk​≥0 and ∑kπk=1\sum_k \pi_k = 1∑k​πk​=1), the function α↦Gβ(x,α,π)\alpha \mapsto G_\beta(x,\alpha,\pi)α↦Gβ​(x,α,π) is convex on R\mathbb RR;
  2. for every α∈R\alpha \in \mathbb Rα∈R, the function π↦Gβ(x,α,π)\pi \mapsto G_\beta(x,\alpha,\pi)π↦Gβ​(x,α,π) is an affine function on RS\mathbb R^SRS:
Gβ(x,α,π)=α+∑k=1Sπk ck(α),ck(α)=[f(x,y[k])−α]+1−β.G_\beta(x,\alpha,\pi) = \alpha + \sum_{k=1}^S \pi_k\, c_k(\alpha), \qquad c_k(\alpha) = \frac{[f(x,y_{[k]})-\alpha]^+}{1-\beta}.Gβ​(x,α,π)=α+k=1∑S​πk​ck​(α),ck​(α)=1−β[f(x,y[k]​)−α]+​.

These are the two structural properties under which the minimax lemma (Lemma 1) applies to GβG_\betaGβ​ on a product of an interval of thresholds α\alphaα and the ambiguity set of distributions π\piπ; affinity in π\piπ gives in particular the concavity in π\piπ that the lemma needs.

Formalization Note Convexity in α\alphaα is stated as ConvexOn ℝ Set.univ for each π\piπ in stdSimplex ℝ (Fin S); only π≥0\pi \ge 0π≥0 and β<1\beta < 1β<1 matter for it. Affinity in π\piπ is stated as the existence of an affine map (Fin S → ℝ) →ᵃ[ℝ] ℝ that agrees with Gβ(x,α,⋅)G_\beta(x,\alpha,\cdot)Gβ​(x,α,⋅) everywhere.

Preamble
import Mathlib
import Definitions.Def_WorstCaseCVaR_Discrete_Setting
Formal statement
namespace WorstCaseCVaR.Discrete

/-- Proof of Theorem 2, Zhu & Fukushima (2009), p. 1167: for fixed `x`, `G_β(x, α, π)` is convex
in `α` (for each probability vector `π`) and affine in `π` (for each `α`). -/
theorem convex_affine_G {n m S : ℕ} (f : (Fin n → ℝ) → (Fin m → ℝ) → ℝ)
    (ys : Fin S → Fin m → ℝ) (x : Fin n → ℝ) (β : ℝ) (hβ0 : 0 < β) (hβ1 : β < 1) :
    (∀ π ∈ stdSimplex ℝ (Fin S), ConvexOn ℝ Set.univ (fun α : ℝ => G f ys β x α π)) ∧
    (∀ α : ℝ, ∃ g : (Fin S → ℝ) →ᵃ[ℝ] ℝ, ∀ π : Fin S → ℝ, G f ys β x α π = g π) := by sorry

end WorstCaseCVaR.Discrete
Source
Zhu & Fukushima, Worst-Case Conditional Value-at-Risk with Application to Robust Portfolio Management, Oper. Res. 57(5), 2009, p. 1167, proof of Theorem 2 (convexity in α cited from Rockafellar and Uryasev 2002)
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