Proof of Proposition 2 — a feasible point of (24)–(30) projects to a feasible point of (16)–(20)
ProvedWorstCaseCVaR.Box.feasible_24_to_16cvarlinear-programmingp2o-batch-pfp2bp2o-gran-per-chapterp2o-plan-paperp2o-v1robust-optimization
Let , scenarios , a set , , and box data be given. If is feasible to (24)–(30), then is feasible to (16)–(20) with ; in particular
Together with the reverse transfer, this shows the two problems have the same optimal value, which is the content of Proposition 2.
Formalization Note Constraint (18) is in its "for all " form (see the definition of the problems). No convexity of and no structure on is assumed.
Preamble
import Mathlib import Definitions.Def_WorstCaseCVaR_Box_Problems open Matrix
Formal statement
namespace WorstCaseCVaR.Box
/-- Proof of Proposition 2, p. 1167: if `(x, u, z, ξ, ω, α, θ)` is feasible to (24)–(30),
then `(x, u, α, θ)` is feasible to (16)–(20) with `𝒫_π = 𝒫_π^B`. -/
theorem feasible_24_to_16 {n m S : ℕ} (f : (Fin n → ℝ) → (Fin m → ℝ) → ℝ)
(ys : Fin S → Fin m → ℝ) (𝒳 : Set (Fin n → ℝ)) (β : ℝ) (hβ0 : 0 < β) (hβ1 : β < 1)
(π0 ηlo ηhi : Fin S → ℝ) :
∀ q ∈ feas24 f ys 𝒳 β π0 ηlo ηhi, q.toVar16 ∈ feas16 f ys 𝒳 β π0 ηlo ηhi := by sorry
end WorstCaseCVaR.Box
Source
Zhu & Fukushima, Worst-Case Conditional Value-at-Risk with Application to Robust Portfolio Management, Oper. Res. 57(5), 2009, p. 1167, proof of Proposition 2
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.