Proof of Theorem 1 — is a nonempty closed bounded interval
ProvedWorstCaseCVaR.Mixture.argmin_intervalconvex-analysiscvarp2o-batch-pfp2bp2o-gran-per-chapterp2o-plan-paperp2o-v1
Let be a probability distribution on , let be integrable under , and let . Then the set of minimizers of the Rockafellar–Uryasev function over is a nonempty, closed, bounded interval: there are with
Applied to each likelihood distribution , this gives the intervals used to localize the minimization in the proof of Theorem 1.
Formalization Note Both bounds on matter: at the minimizer set can be a half-line.
Preamble
import Mathlib import Definitions.Def_WorstCaseCVaR_Mixture_Setting open MeasureTheory
Formal statement
namespace WorstCaseCVaR.Mixture
theorem argmin_interval {m n : ℕ} (P : Measure (Fin m → ℝ)) [IsProbabilityMeasure P]
(f : (Fin n → ℝ) → (Fin m → ℝ) → ℝ) (β : ℝ) (hβ0 : 0 < β) (hβ1 : β < 1)
(x : Fin n → ℝ) (hf : Integrable (f x) P) :
∃ a b : ℝ, a ≤ b ∧ {α : ℝ | IsMinOn (ruFun P f β x) Set.univ α} = Set.Icc a b := by sorry
end WorstCaseCVaR.Mixture
Source
Zhu & Fukushima, Worst-Case Conditional Value-at-Risk with Application to Robust Portfolio Management, Oper. Res. 57(5), 2009, p. 1166, proof of Theorem 1 (citing Rockafellar and Uryasev 2000, 2002)
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.